View in EDS HTML Full Text PDF Full Text

Effective Writing Instruction for Students in Grades 6 to 12: A Best Evidence Meta-Analysis

Saved in:
Bibliographic Details
Title: Effective Writing Instruction for Students in Grades 6 to 12: A Best Evidence Meta-Analysis
Language: English
Authors: Steve Graham (ORCID 0000-0002-6702-5865), Yucheng Cao, Young-Suk Grace Kim, Joongwon Lee, Tamara Tate, Penelope Collins, Minkyung Cho, Youngsun Moon, Huy Quoc Chung, Carol Booth Olson
Source: Reading and Writing: An Interdisciplinary Journal. 2025 38(4):1-46.
Availability: Springer. Available from: Springer Nature. One New York Plaza, Suite 4600, New York, NY 10004. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-460-1700; e-mail: customerservice@springernature.com; Web site: https://link.springer.com/
Peer Reviewed: Y
Page Count: 46
Publication Date: 2025
Sponsoring Agency: Institute of Education Sciences (ED)
Contract Number: R305C190007
Document Type: Journal Articles
Information Analyses
Reports - Research
Education Level: Secondary Education
Descriptors: Writing Instruction, Secondary Education, Instructional Effectiveness, Writing Skills, Educational Research, Research Methodology, Meta Analysis
DOI: 10.1007/s11145-024-10539-2
ISSN: 0922-4777
1573-0905
Abstract: The current best evidence meta-analysis reanalyzed the data from a meta-analysis by Graham et al. (J Educ Psychol 115:1004-1027, 2023). This meta-analysis and the prior one examined if teaching writing improved the writing of students in Grades 6 to 12, examining effects from writing intervention studies employing experimental and quasi-experimental designs (with pretests). In contrast to the prior meta-analysis, we eliminated all N of 1 treatment/control comparisons, studies with an attrition rate over 20%, studies that did not control for teacher effects, and studies that did not contain at least one reliable writing measure (0.70 or greater). Any writing outcome that was not reliable was also eliminated. Across 148 independent treatment/control comparisons, yielding 1,076 writing effect sizes (ESs) involving 22,838 students, teaching writing resulted in a positive and statistically detectable impact on students' writing (ES = 0.38). Further, six of the 10 writing treatments tested in four or more independent comparisons improved students' performance. This included the process approach to writing (0.75), strategy instruction (0.59), transcription instruction (0.54), feedback (0.30), pre-writing activities (0.32), and peer assistance (0.59). In addition, the Self-Regulated Strategy Development model for teaching writing strategies yielded a statistically significant ES of 0.84, whereas other approaches to teaching writing strategies resulted in a statistically significant ES of 0.51. The findings from this meta-analysis and the Graham et al. (2023) review which included studies that were methodologically weaker were compared. Implications for practice, research, and theory are presented.
Abstractor: As Provided
IES Funded: Yes
Entry Date: 2025
Accession Number: EJ1468906
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwFyMvIRTebwr-aWNPl0OMdZAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDLnFopxDCm4BDg7CpQIBEICBm6FyDb3gQYxlOuqzdzGAZ7HLGh71fMEc9LTwAGtS0sxy4M6mybTxVprWFQonU_ccR60-a8gPEdBC-OZ3Rt02mVo97sPAGeqzp6-oC0xWHK-CUcDE8S-PJ-PTDBJgIh2Fc3uYdStgDNDrXBuwsEMFCoQT-tX8mxUmF5bjcU7UK2JnG94qc7ApAEOeydvim-3QJd7WRu01DTqGHIMi
Text:
  Availability: 1
  Value: <anid>AN0184705652;2ap01apr.25;2025Apr28.03:02;v2.2.500</anid> <title id="AN0184705652-1">Effective writing instruction for students in grades 6 to 12: a best evidence meta-analysis </title> <p>The current best evidence meta-analysis reanalyzed the data from a meta-analysis by Graham et al. (J Educ Psychol 115:1004–1027, 2023). This meta-analysis and the prior one examined if teaching writing improved the writing of students in Grades 6 to 12, examining effects from writing intervention studies employing experimental and quasi-experimental designs (with pretests). In contrast to the prior meta-analysis, we eliminated all N of 1 treatment/control comparisons, studies with an attrition rate over 20%, studies that did not control for teacher effects, and studies that did not contain at least one reliable writing measure (0.70 or greater). Any writing outcome that was not reliable was also eliminated. Across 148 independent treatment/control comparisons, yielding 1,076 writing effect sizes (ESs) involving 22,838 students, teaching writing resulted in a positive and statistically detectable impact on students' writing (ES = 0.38). Further, six of the 10 writing treatments tested in four or more independent comparisons improved students' performance. This included the process approach to writing (0.75), strategy instruction (0.59), transcription instruction (0.54), feedback (0.30), pre-writing activities (0.32), and peer assistance (0.59). In addition, the Self-Regulated Strategy Development model for teaching writing strategies yielded a statistically significant ES of 0.84, whereas other approaches to teaching writing strategies resulted in a statistically significant ES of 0.51. The findings from this meta-analysis and the Graham et al. (2023) review which included studies that were methodologically weaker were compared. Implications for practice, research, and theory are presented.</p> <p>Keywords: Writing; Instruction; Teaching; Middle school; High school; Meta-analysis</p> <p>Copyright comment Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.</p> <hd id="AN0184705652-2">Introduction</hd> <p>Writing is an essential and fundamental skill that is essential to adolescent students' success at school. Secondary students use writing to enhance their understanding of materials read or presented in class (Bangert-Drowns et al., [<reflink idref="bib6" id="ref1">6</reflink>]; Graham & Hebert, [<reflink idref="bib62" id="ref2">62</reflink>]; Graham et al., [<reflink idref="bib65" id="ref3">65</reflink>]). By engaging in writing activities, such as exams and reports, students demonstrate their mastery of content material (Graham, [<reflink idref="bib56" id="ref4">56</reflink>]). Writing even enhances adolescent students' reading capabilities, as teaching them to write enhances multiple reading skills (Graham & Hebert, [<reflink idref="bib62" id="ref5">62</reflink>]).</p> <p>While students learn to write in multiple contexts and communities (Graham, [<reflink idref="bib57" id="ref6">57</reflink>]), a basic responsibility of schools is to help them become competent writers. This responsibility begins at the very start of the schooling process (Tolchinsky, [<reflink idref="bib156" id="ref7">156</reflink>]), and it continues throughout the secondary grades. It is expected that the writing capabilities acquired in the elementary grades will be extended in middle and high school, and students will become increasingly adept at using writing as a tool for analyzing and learning content in subject-matter classes (Hillocks, [<reflink idref="bib77" id="ref8">77</reflink>]). Unfortunately, many secondary schools are not adequately meeting their responsibilities to teach writing. For instance, two out of every three secondary students' scores on the last national assessment in the United States revealed that secondary students had not mastered the writing capabilities needed for grade-level success (National Center for Educational Statistics, [<reflink idref="bib118" id="ref9">118</reflink>]). Such findings are not limited to the United States, as they are common in countries around the world (Graham & Rijlaarsdam, [<reflink idref="bib68" id="ref10">68</reflink>]).</p> <p>If teachers in secondary schools are to do a better job of teaching writing to their students, they need access to teaching practices with a track record of proven success. This is especially important because concerns have been raised in many countries about the quality of writing instruction secondary students receive (e.g. Applebee & Langer, [<reflink idref="bib4" id="ref11">4</reflink>]; Drew et al., [<reflink idref="bib41" id="ref12">41</reflink>]; Graham, [<reflink idref="bib56" id="ref13">56</reflink>]). A useful approach to identifying potentially effective writing practices is the systematic review of intervention research studies that test the impact of specific writing procedures and practices. Meta-analysis provides one approach for identifying effective writing practices for secondary students. This review procedure is used to summarize the direction and magnitude of the effects obtained by a set of research investigations examining the same basic phenomena (Lipsey & Wilson, [<reflink idref="bib105" id="ref14">105</reflink>]).</p> <p>Most prior meta-analyses of writing interventions have focused just on the elementary grades (e.g., Graham et al., [<reflink idref="bib66" id="ref15">66</reflink>]; Koster et al., [<reflink idref="bib98" id="ref16">98</reflink>]) or a combination of elementary and secondary grades (e.g., Bangert-Drowns, [<reflink idref="bib5" id="ref17">5</reflink>]; Goldberg et al., [<reflink idref="bib53" id="ref18">53</reflink>]; Graham & Harris, [<reflink idref="bib59" id="ref19">59</reflink>]; Graham et al., [<reflink idref="bib63" id="ref20">63</reflink>]; Hillocks, [<reflink idref="bib78" id="ref21">78</reflink>]; Morphy & Graham, [<reflink idref="bib116" id="ref22">116</reflink>]; Rogers & Graham, [<reflink idref="bib138" id="ref23">138</reflink>]), with little emphasis on determining the separate effects of specific writing interventions with just secondary students. Even the meta-analysis by Graham and Perin ([<reflink idref="bib67" id="ref24">67</reflink>]), which presumably was aimed at adolescent students, included studies conducted with children in Grades 4 and 5. This review did provide separate effects for secondary students, but this was limited to just one writing intervention: the process approach to writing (see also Sandmel & Graham, [<reflink idref="bib144" id="ref25">144</reflink>]).</p> <p>A more recent meta-analysis of writing interventions by Graham and colleagues ([<reflink idref="bib64" id="ref26">64</reflink>]) focused just on students in the secondary grades (Grades 6 to 12). This review included 406 independent writing treatment/control comparisons from published and unpublished studies involving 52,529 secondary students. All studies applied an experimental or quasi-experimental design (with pretests). When all writing outcomes were considered collectively, teaching and practicing writing yielded a positive and statistically detectable effect size (ES) on students' writing (0.47). When just specific writing outcomes were analyzed, writing instruction statistically enhanced writing quality (ES = 0.46); ideation (ES = 0.46); text organization (ES = 0.51); inclusion of genre elements (ES = 0.60); voice (ES = 0.30); sentence construction skills (ES = 0.51); handwriting, spelling, and typing (ES = 0.49); planning, revising, and writing self-regulation processes (ES = 0.82); knowledge about writing (ES = 0.94); and writing motivational beliefs (ES = 0.13).</p> <p>Graham and associates ([<reflink idref="bib64" id="ref27">64</reflink>]) also examined if 23 specific writing interventions positively impacted secondary students' writing outcomes. Statistically detectable positive effects were obtained for the following 17 writing practices: comprehensive writing programs (ES = 0.47); writing strategy instruction (ES = 0.76); use of digital writing tools (ES = 0.31); teaching handwriting, spelling, and/or typing (ES = 0.71); computer-assisted instruction (ES = 0.32); teaching critical/creative thinking skills for writing (ES = 0.27); emulating good models of writing (ES = 0.46); feedback (ES = 0.34); goal setting (ES = 0.44); pre-writing activities (ES = 0.49); grammar instruction (ES = 0.77); sentence instruction (ES = 0.73); inquiry (ES = 0.92); observing writers/readers (ES = 0.41); peer assistance (ES = 0.41); summarization instruction (ES = 0.49); and text structure instruction (ES = 0.39). The following six writing practices did not statistically enhance writing performance: vocabulary instruction, revision instruction, procedural facilitation to enhance writing processes, procedures to enhance motivation, teaching the use of imagery, and increasing writing practice.</p> <p>The analyses undertaken by Graham et al. ([<reflink idref="bib64" id="ref28">64</reflink>]) demonstrated that secondary students' writing outcomes, including their foundational writing skills, processes and strategies students use to write, knowledge of writing, and writing motivational beliefs can be improved through writing instruction. The outcomes from this previous meta-analysis also demonstrated that there are multiple instructional writing practices that teachers can use to help ensure their students achieve positive writing outcomes. Despite these positive effects, there was a considerable variability in the obtained effects across the 406 writing treatment/control comparisons. Some of this variability in ESs was associated with study quality which included: research design (experimental vs quasi-experiment), publication type (journal article vs non-journal article), norm-referenced test (yes or no), treatment fidelity (treatment fidelity established vs not established), N of 1 study (not N of 1 study vs N of 1 study), teacher effects controlled (teacher effects controlled vs not controlled), attrition rate (attrition rate greater than 20% vs less than 20%), and measurement reliability (measurement reliability 0.70 or higher vs less than 0.70).</p> <p>Three of these study quality indicators, publication type, norm-referenced test, and not N of 1 study, accounted for statistically significant unique variability in ESs for writing. The effects for journal articles were higher than the effects for non-journal articles, as journal articles had an ES that was 0.21 larger than the effect for non-journal articles. This was likely because reviewers are more likely to recommend the publication of studies that produce more positive outcomes (Hopewell et al., [<reflink idref="bib87" id="ref29">87</reflink>]). The outcome for norm-referenced tests was 0.15 of a SD lower than it was for tests that were not norm-referenced. This was likely because tests that are not norm-referenced may be over aligned with the writing treatment (Slavin & Madden, [<reflink idref="bib147" id="ref30">147</reflink>]). Finally, a study that is a N of 1 investigation includes a treatment, control, or both conditions where all students are taught together by one instructor. The outcome for such studies was 0.20 of a SD higher than it was for studies that were not N of 1. An N of 1 study increases the likelihood that obtained ESs are due to teacher effects (e.g., one condition has a particularly motivating teacher and the other does not).</p> <p>The finding that study quality indicators were correlated with study outcomes in the Graham et al.'s ([<reflink idref="bib64" id="ref31">64</reflink>]) meta-analysis raised the possibility of biases in the findings (see Cheung & Slavin, [<reflink idref="bib17" id="ref32">17</reflink>]). In other words, obtained outcomes may be larger or smaller when such biases are eliminated. When biases related to study quality are not explored, there is the possibility that the findings from the resulting meta-analyses may mislead teachers, policy makers, and researchers about the effectiveness of an intervention. To address this, we reanalyzed the writing data from Graham et al. ([<reflink idref="bib64" id="ref33">64</reflink>]) in the present review, eliminating all N of 1 treatment/control comparisons, studies with an attrition rate over 20%, studies that did not control for teacher effects, and studies that did not contain at least one reliable writing measure (0.70 or greater). We also eliminated effects for any writing outcome that did not evidence reliability. For example, if a study had three writing outcomes and just one them was established as reliable, the study was retained but an effect was only calculated for the reliable outcome.</p> <p>The reason for eliminating studies that were N of 1 was because such investigations can be biased due to teacher effects. Because studies that are not N of 1 can also contain biasing teacher effects (e.g., treatment teachers were volunteers), we also eliminated treatment/control comparisons that did not put into place procedures to control for such effects (i.e., instructors randomly assigned to conditions, or all instructors taught each condition). Studies with attrition greater than 20% were further eliminated because attrition at this level can change the outcomes of a study (Shulz & Grimes, [<reflink idref="bib145" id="ref34">145</reflink>]). Lastly, we eliminated any writing outcome where reliability was not established (0.70 or higher) because outcomes from such assessments may be inconsistent with outcomes from reliable assessments (Nunnally, [<reflink idref="bib120" id="ref35">120</reflink>]).</p> <p>It is important to note that we did not eliminate quasi-experimental studies, treatment/control comparisons where treatment fidelity was not established, non-journal publications, or outcomes that were not based on norm-referenced tests. While quasi-experiments may produce larger effects than experiments because selective factors may work in favor of the former (Cheung & Slavin, [<reflink idref="bib17" id="ref36">17</reflink>]), we did not exclude them because the quasi-experiments in this review controlled for pretest differences (minimizing the impact of biasing selective factors) and quasi-experiments do not necessarily produce different effects than experiments when assessing the impact of writing treatments (Aiken et al., [<reflink idref="bib2" id="ref37">2</reflink>]). Studies where treatment fidelity was not established were also not eliminated for the following reason. Establishing treatment fidelity is a relatively new concept in writing intervention research (Graham & Harris, [<reflink idref="bib61" id="ref38">61</reflink>]). Thus, many previously tested writing treatments may have been implemented with fidelity, but this cannot be adequately determined because this has rarely been assessed (only 31% of the studies in Graham et al., [<reflink idref="bib64" id="ref39">64</reflink>] assessed treatment fidelity). Non-journal publications were not eliminated because this would overestimate the effects of writing treatments as demonstrated in Graham et al. ([<reflink idref="bib64" id="ref40">64</reflink>]). Additionally, we did not eliminate ESs for writing outcomes that were not norm-referenced assessments because 93% of all ESs from Graham et al. ([<reflink idref="bib64" id="ref41">64</reflink>]) would have been eliminated, leaving us with virtually nothing to analyze.</p> <hd id="AN0184705652-3">Theoretical underpinnings</hd> <p></p> <hd id="AN0184705652-4">Writing</hd> <p>The Writer(s)-within-Community model of writing (WWC; Graham, [<reflink idref="bib57" id="ref42">57</reflink>]) served as the theoretical foundation for the Graham et al. ([<reflink idref="bib64" id="ref43">64</reflink>]) meta-analysis as well as this best evidence meta-analysis of writing interventions for secondary students. The basic assumption underlying the WWC is writing and the teaching of writing are simultaneously and interactively shaped by the context in which each occurs and the cognitive capabilities and resources of the writers in these contexts. Within a writing community, the teaching of writing is determined and bound by its purposes, including writing goals, values, motives, audiences, established norms, social practices, writing supports, writing stance/identity, available writing tools, typified writing practices, types of writing products produced, experiences and resources of classroom members (teachers and students in classrooms), physical and social environment, and collective history of the community. What is written and how writing is taught is also shaped by institutional, social, cultural, political, and historical factors operating outside the writing community.</p> <p>Writing and teaching writing within a secondary classroom is accomplished by teachers and students, and through the cognitive capabilities and resources they possess (Graham, [<reflink idref="bib57" id="ref44">57</reflink>]). Within the WWC model, students deliberately and consciously exert executive control over the composing process by determining their own writing goals (usually in concert with teacher and classroom goals). They achieve these goals by activating, orchestrating, and adjusting as needed relevant knowledge and beliefs held in long-term memory, available writing tools and external resources (e.g., information from print or digital materials), and the production processes used to create writing. The use of these cognitive and external resources may be altered by emotional responses to writing, personality traits, and physiological states. These cognitive resources and processes (executive control mechanisms, long-term memory resources, production processes, and emotional, psychological, and physical modulators) not only shape and constrain the writing actions of students in the classroom, but the instructional actions of their teachers as well (Graham, [<reflink idref="bib64" id="ref45">64</reflink>]).</p> <p>According to the WWC model, the control processes students and teachers use when writing and teaching writing, respectively, include directing and maintaining attention, working memory, and executive control procedures including setting goals for writing, devising plans or using preexisting schemas to accomplish these objectives, and monitoring and adjusting goals and plans as needed. Long-term memory (LTM) sources include specialized knowledge about writing (e.g., knowledge of spelling; handwriting; typing; keyboarding; punctuation; capitalization; grammar and sentence instruction; purposes and features of different types of text; qualities of good writing; schemas and strategies for planning, drafting, and revising text; and tools for writing) as well as knowledge about the writing topic, intended audiences, oral language (which forms the foundation for writing), and reading (writers use reading to gather information for writing and when they evaluate and revise their writing). LTM also includes a variety of beliefs that can power and limit students' writing efforts, including efficacy for writing, attitudes/interest in writing, value and utility placed on writing, motives for writing, reasons for writing success/failure, and students' identities as writers. Writers also use specific processes to produce text. These production processes include: conceptualization (goals, plans, and mental representation of the writing task), ideation (acquiring writing content from LTM or external sources), translation (turning ideas into acceptable phrases and sentences), transcription (transcribing phrases and sentences into written text), and reconceptualization (rethinking and revising plans and text).</p> <p>In Graham et al. ([<reflink idref="bib64" id="ref46">64</reflink>]) and the current best evidence meta-analysis, the WWC model (Graham, [<reflink idref="bib57" id="ref47">57</reflink>]) provided a roadmap for reducing the multitude of different writing outcomes in reviewed studies into a more manageable set of 10 writing constructs (see Coding of Study Features in the Method section). For instance, the skills involved in constructing sentences (e.g., sentence fluency, syntax, vocabulary, and writing conventions) were grouped together to represent the WWC production process of translation. Likewise, assessments of attitude towards writing, writing self-efficacy, and other writing beliefs were combined to represent the construct of writing motivational beliefs. Moreover, the WWC model provided us with a framework for categorizing different writing treatments and a framework for considering what cognitive processes or LTM resources an effective writing treatment was designed to impact (see Table 1). For example, teaching handwriting, spelling, and/or typing was categorized as transcription instruction. Teaching writing strategies was categorized as conceptualization and reconceptualization production processes as well as schemas for guiding executive control.</p> <p>Table 1 Definition for coded writing treatments and cognitive processes from WWC model impacted (in parentheses)</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Writing treatment</p></th><th align="left"><p>Definition</p></th></tr></thead><tbody><tr><td align="left"><p>Process approach to writing</p></td><td align="left"><p>The process approach to writing (sometimes referred to as writers' workshop) involves extended opportunities for writing; writing for real audiences; engaging in cycles of planning, translating, and reviewing; personal responsibility and ownership of writing projects; high levels of student interactions; creation of a supportive writing classroom; self-reflection; personalized individual assistance and instruction</p></td></tr><tr><td align="left"><p>Strategy instruction</p></td><td align="left"><p>Strategy instruction involves teaching students strategies for planning, revising, self-assessing, and/or editing text. An instructional approach is not identified as strategy instruction unless the procedure is modeled and students practice applying it at least two times. If writing strategies were taught using the Self-Regulated Strategy Development model (Harris & Graham, <xref ref-type="bibr" rid="bibr71">1992</xref>), this was coded. This approach is a special form of strategy instruction that emphasizes teaching students the knowledge, skills, and self-regulation procedures needed to use target strategies successfully. (conceptualization, reconceptualization, executive control)</p></td></tr><tr><td align="left"><p>Digital writing</p></td><td align="left"><p>This involved the use of digital tools for writing, including word processing; word processing plus programs with additional features such as planning or evaluation aids, digital game elements, and developing a context in which writing occurs; as well as speech-to-text synthesis. To qualify as digital writing, students had to use these tools over multiple sessions. (transcription, conceptualization, reconceptualization, ideation, motivational beliefs)</p></td></tr><tr><td align="left"><p>Transcription instruction</p></td><td align="left"><p>Included teaching spelling, handwriting, or keyboarding. (transcription)</p></td></tr><tr><td align="left"><p>Computer-assisted instruction</p></td><td align="left"><p>Using a computer to teach writing skills, processes, or knowledge. (all writing production processes)</p></td></tr><tr><td align="left"><p>Feedback</p></td><td align="left"><p>Students received feedback about the adequacy of their text or progress in learning writing skills or processes. Feedback was further coded as originating from teachers, peers, computers, or self. (reconceptualization)</p></td></tr><tr><td align="left"><p>Pre-writing activities</p></td><td align="left"><p>Use of activities such as using semantic webs, mind maps, brainstorming, searching the web for ideas, reading text for ideas, dialoguing with others designed to help students generate or organize ideas for writing. (ideation)</p></td></tr><tr><td align="left"><p>Peer assistance</p></td><td align="left"><p>Students working together to plan, draft, edit, and/or revise their compositions as well as learning via peer-tutoring, cooperative learning, or students working together. (all writing production processes)</p></td></tr><tr><td align="left"><p>Text structure instruction</p></td><td align="left"><p>Explicitly teaching students knowledge about the purpose and/or structure of specific types of text, such as stories or persuasive essays. (writing knowledge)</p></td></tr><tr><td align="left"><p>Increased writing</p></td><td align="left"><p>Increasing how much students write. (all writing production processes)</p></td></tr></tbody></table> </ephtml> </p> <hd id="AN0184705652-5">Research questions and predictions</hd> <p>This review was designed to answer the following questions for the studies that met our best evidence criteria described earlier.</p> <p></p> <ulist> <item> Does teaching writing improve writing outcomes collectively, as well as specific outcomes of writing quality, ideation, elements, organization, voice, writing process, translation skills, transcription skills, and writing motivational beliefs?</item> <p></p> <item> Are effects for writing outcomes moderated by the quality of studies (publication type, treatment fidelity, and norm-referenced test) and specific study features (grade level, typical vs. less capable writers, and type of control condition)?</item> <p></p> <item> What specific writing treatments improve writing outcomes?</item> </ulist> <p>To answer Research Question 1, we first computed an ES for all writing outcomes collectively. ESs were also calculated for nine writing constructs identified earlier from the WWC model (Graham, [<reflink idref="bib57" id="ref48">57</reflink>]; e.g., writing quality, writing output, writing process, and writing knowledge). It is not only important to determine the impact of writing treatments overall, but to determine if specific aspects of writing are improved as well. This can identify which aspects of writing our best evidence treatment/control comparisons did or did not improve. As with Graham et al. ([<reflink idref="bib64" id="ref49">64</reflink>]), we anticipated that collective and all specific writing outcomes would be positively impacted, but not as strongly as in the previous meta-analysis because higher quality studies tend to produce smaller effects (Cheung & Slavin, [<reflink idref="bib17" id="ref50">17</reflink>]).</p> <p>For Research Question 2, two separate moderator analyses were conducted. One moderator analysis examined if the magnitude of effects was related to three study quality indicators that were not used to eliminate studies from Graham et al. ([<reflink idref="bib64" id="ref51">64</reflink>]): publication type, treatment fidelity, and norm-referenced test. As in the previous meta-analysis by Graham et al. ([<reflink idref="bib64" id="ref52">64</reflink>]), we again anticipated these study quality predictors would be associated with variability in magnitude of writing effects.</p> <p>A second moderator analysis focused on whether the magnitude of ESs for all writing outcomes collectively differed by three study features (grade-level, typical vs. less capable writers, and type of control condition). These study features were selected as moderators because (a) studies with younger students may generate larger effects than studies with older students because the former tend to make greater academic gains in a year than the latter (Bloom et al., [<reflink idref="bib12" id="ref53">12</reflink>]); (b) typically developing students tend to make greater academic gains during a year than less capable ones (Swanson et al., [<reflink idref="bib150" id="ref54">150</reflink>]); and (c) the magnitude of effects for writing treatments may differ depending on the type of control condition to which it is compared (Pressley et al., [<reflink idref="bib130" id="ref55">130</reflink>]).</p> <p>For Research Question 3, we computed an ES for 10 writing treatments (e.g., pre-writing activities, computer-assisted instruction). An ES for a writing treatment was only calculated if there were at least four or more independent treatment/control comparisons (a criterion common in most prior writing meta-analyses; Graham & Perin, [<reflink idref="bib67" id="ref56">67</reflink>]). We anticipated that most of the writing treatments tested would produce statistically significant and positive effects, as was the case in Graham et al, ([<reflink idref="bib64" id="ref57">64</reflink>]), but effect sizes would generally be smaller than they were in the prior meta-analysis because stronger studies tend to produce smaller effects (Cheung & Slavin, [<reflink idref="bib17" id="ref58">17</reflink>]).</p> <p>Like Graham et al. ([<reflink idref="bib64" id="ref59">64</reflink>]), separate ESs were computed for specific approaches for two of the writing treatments: strategy instruction and digital writing mode. For strategy instruction, ESs were computed for studies applying the Self-Regulated Strategy Development model (SRSD; Harris & Graham, [<reflink idref="bib71" id="ref60">71</reflink>]) and for studies not applying this approach to strategy instruction. In contrast to other approaches to writing strategy instructional approaches, SRSD also teaches students self-regulation procedures for regulating learned strategies, the writing process, and writing behaviors as well as the knowledge needed to use targeted writing strategies and self-regulation procedures. In addition, two separate ESs were calculated for the writing treatment using digital writing tools. This included calculating an ES for using word processing as well as an ES for word processing plus programs. Word processing plus programs included features that are not commonly included with basic word processing programs, such as software to facilitate planning when writing, game features to make writing more motivating, and computer-provided feedback (see Morphy & Graham, [<reflink idref="bib116" id="ref61">116</reflink>]).</p> <hd id="AN0184705652-6">Method</hd> <p></p> <hd id="AN0184705652-7">Inclusion criteria</hd> <p>Each study in this review met the following criteria: (<reflink idref="bib1" id="ref62">1</reflink>) students were in Grades 6 to 12, (<reflink idref="bib2" id="ref63">2</reflink>) students attended public or private schools, (<reflink idref="bib3" id="ref64">3</reflink>) students in the writing treatment wrote text, were taught writing, or both (50% or more of instruction was devoted to teaching writing), (<reflink idref="bib4" id="ref65">4</reflink>) the study applied an experimental or quasi-experimental design (with pretest), (<reflink idref="bib5" id="ref66">5</reflink>) some aspect of writing was assessed (e.g., written products, writing skills, writing processes, writing motivational beliefs, or writing); (<reflink idref="bib6" id="ref67">6</reflink>) the study was written in English; and (<reflink idref="bib7" id="ref68">7</reflink>) data needed to calculate ESs were provided or available from authors.</p> <p>Studies were excluded when students attended special schools or were educated in residential settings (e.g., school for students identified as deaf or hard of hearing, residential setting for students with behavioral difficulties). Although writing instruction is important for these students, the purpose of this meta-analysis was to draw best-evidence recommendations for teaching writing in general education settings. Studies were further excluded if there were fewer than 10 students either in the treatment or control conditions. Studies with a small number of students have a larger margin of error than studies with a larger sample of students. Studies were also excluded if they assessed some aspect of test administration (e.g., intended audience); compared writing digitally to writing by hand without practice applying these modes of writing over time; or assessed the effectiveness of a writing treatment with single participant designs, designs where students served as their own controls, regression discontinuity designs, and ex post facto designs. Studies manipulating test administration and assessing different modes of writing at a single point in time do not involve teaching writing, whereas ESs from the four excluded designs (e.g., ex post facto) will not yield comparable metrics to ESs from experimental and quasi-experimental designs (Lipsey & Wilson, [<reflink idref="bib105" id="ref69">105</reflink>]). Lastly, studies were excluded if students in the treatment condition were provided instruction in one year and control students were from classes in preceding or subsequent school years. Studies were not classified as experiments if randomization was done by school administrators before the school year began. This is because in our experience, adjustments in class rosters are still made after randomization due to interactions between principals, teachers, and parents.</p> <p>The inclusion/exclusion criteria specified above were used to identify possible studies for Graham et al. ([<reflink idref="bib64" id="ref70">64</reflink>]). Because the purpose of the current review was to only to examine investigations from Graham et al. ([<reflink idref="bib64" id="ref71">64</reflink>]) that met our additional criteria for best evidence, we eliminated studies that were N of 1, had an attrition rate over 20%, did not control for teacher effects, and did not contain at least one reliable writing measure (0.70 or greater). Moreover, any writing measure that was not reliable for an included study was not included in our analyses.</p> <hd id="AN0184705652-8">Search procedures and selection of studies</hd> <p>The following strategies were applied by Graham et al. ([<reflink idref="bib64" id="ref72">64</reflink>]) to identify possible studies for their review. Electronic searches, ending September 30, 2021, were conducted in ERIC, PsycINFO, Linguistic and Language Behavior Abstracts, and ProQuest Dissertations & Theses Global. The following search terms (including widows) were used in each search: (writing) AND (instruct* OR treat* OR teach* OR handwrit* OR spell* OR grammar OR vocabulary OR "strategy instruction" OR "planning instruction" OR "revising instruction" OR sentence-combin* OR "sentence instruction" OR "process approach" OR "computer-assisted instruct*" OR "word process*" OR "writ* workshop" OR "national writ* project" OR "peer collaborat*" OR "collaborat* writ*" OR "automated essay" OR "goal set*" OR model OR "self-regulated strategy development" OR feedback OR usage OR mechanics OR "word processor" OR technology OR "on-line" OR "motivation and instruction" OR "free writing" OR "additional writing" OR "extra writing" OR "peer-planning" OR "peer-revising" OR "self-evaluation" OR "self-monitoring" OR "evaluative scales" OR "automated writing evaluation" OR "mentor text" OR "formative assessment") AND ("Empirical Study" OR "Quantitative Study" OR "Longitudinal Study" OR "Followup Study" OR "Experimental Replication" OR "true experiment*" OR "quasi experiment*" OR "randomized control" OR "randomized cluster").</p> <p>Sixteen journals were also hand searched from first date of publication to September 30, 2021. These journals included <emph>American Educational Research Journal, Assessing Writing</emph>, <emph>British Educational Research Journal, British Journal of Educational Psychology, Journal of Educational Psychology, Journal of Educational Research, Journal of Literacy Research, Journal of Research in Reading, Journal of Research in Teaching English, Journal of Research on Educational Effectiveness, Journal of Writing Research, Learning & Instruction, Reading & Writing, Reading Research Quarterly, Scientific Studies of Reading,</emph> and <emph>Written Communication</emph>. These journals were selected because they commonly publish writing studies.</p> <p>Twenty-four prior meta-analyses investigating the effects of writing instruction were consulted to identify possible studies (e.g., Graham & Perin, [<reflink idref="bib67" id="ref73">67</reflink>]; Graham et al., [<reflink idref="bib63" id="ref74">63</reflink>]; Hillocks, [<reflink idref="bib78" id="ref75">78</reflink>]). Additionally, 35 authors who conduct writing intervention research were asked to share relevant studies and citations (e.g., Rui Alves, Raquel Fidalgo, Teresa Limpo, Charles MacArthur, Linda Mason, Deborah Myhill, George Newell, and Gert Rijlaarsdam). References in documents included in the meta-analysis were examined to identify other pertinent studies. The Prisma chart for this search process can be found in Graham et al. ([<reflink idref="bib64" id="ref76">64</reflink>]). The search yielded 28,685 possible documents; 3,549 of these were duplicates. Another 24,020 documents were eliminated by screening abstracts and eliminating documents that did not meet inclusion/exclusion criteria (reliability of screening for 20% of these documents by two judges was 98%). This left 1096 documents. When the full text of these documents was examined, 357 documents were found to be acceptable (reliability of screening for 25% of these documents for two judges was 96%).</p> <p>Five of the 357 documents were not included in the Graham et al.'s ([<reflink idref="bib64" id="ref77">64</reflink>]) meta-analysis because it was not clear which writing interventions were treatment and control in these investigations (e.g., Dunnagan, [<reflink idref="bib42" id="ref78">42</reflink>]). The remaining 352 documents yielded 406 independent treatment/control comparisons. When we eliminated studies in this current review because they were N of 1, attrition exceeded 20%, teacher effects were not controlled, or no writing measures were reliable, the number of included documents dropped considerably to 128 documents, with 148 independent treatment/control comparisons.</p> <hd id="AN0184705652-9">Coding of study features</hd> <p>In Graham et al. ([<reflink idref="bib64" id="ref79">64</reflink>]), all studies in the 352 documents were coded for the following quality indicators: research design (experiment or quasi-experiment), publication type (journal article or other publication), treatment fidelity (established or not established), N of 1 investigation (N of 1 study or not N of 1), teacher effects controlled (controlled [teacher effects controlled by random assignment of instructors to conditions or each instructor teaching all conditions] or not controlled), and attrition greater than 20% (yes or no). In addition, each writing and reading measure were coded to determine if they were reliable (0.70 or greater) and if it was norm-referenced. Since studies in the current review were not N of 1, controlled for teacher effects, had attrition less than 20%, and calculated ESs for only reliable measures, information on these variables are not reported here. In total, we eliminated studies in 144 documents because they were N of 1, 145 documents because they did not control teacher effects, 20 documents because attrition was greater than 20%, and 21 documents because there was no single reliable measure.</p> <p>Graham et al. ([<reflink idref="bib64" id="ref80">64</reflink>]) also coded all studies for gender (male, female), race (Asian, Black, Hispanic, Native, White, and Other), language status (native language learners, second language learners, or a mixture of these two groups of students), socio-economic status (percentage of students receiving free or reduced lunch), grade-level, study location (urban, suburban, rural), country (e.g., Finland), subject area (language arts, science, social studies, math, other), genre (narrative, expository, persuasive, multiple genres, other genres, no writing took place), training of treatment implementers (yes or no), student type (typical students and less capable writers [e.g., the 25th percentile or below, students with special needs]), students composed text when writing (yes or no), instructional time (hours of writing instruction provided), and type of control condition (no instruction, business-as-usual, partial administration of treatment, reading instruction, alternative non-writing treatment other than reading [e.g., math], and alternative writing instruction).</p> <p>All writing measures in a study were described and coded as one of the following 10 outcomes derived earlier from the WWC model (Graham, [<reflink idref="bib57" id="ref81">57</reflink>]): writing quality (e.g., holistic writing scale, analytic writing scale, or total score for norm-referenced writing test), genre elements (e.g., number of schematic elements included in a text, such as the inclusion of characters, setting, and events in narrative writing), ideation (e.g., number of words, sentence, idea-units, or T-units written as well as clarity, substance, and/or richness of ideas conveyed in a written text), organization (e.g., logical progression of ideas in a paper, including the establishment of transitions between paragraphs and sentences as well as coherence), voice, writing process (e.g., planning, revising, perseverance, and self-regulation measures), translation skills (e.g., sentence fluency, syntax, grammar, vocabulary, and convention measures), transcription skills (e.g., spelling, handwriting, and keyboarding measures), knowledge of writing(e.g., knowledge of writing topic, audience, genre, writing task, or conventions), and motivational beliefs (e.g., efficacy, attitudes, and motivation measures such as intrinsic and extrinsic motivation).</p> <p>For the present review, there were three studies where a writing quality score was not provided, but we were able to calculate one from a combination of administered measures assessing ideation, organization, voice, sentence fluency, vocabulary, and conventions. In each, we averaged the outcomes for these attributes (e.g., ideation, organization, sentence fluency) using the Nouri and Greenberg procedure (Cortina & Nouri, [<reflink idref="bib24" id="ref82">24</reflink>]) before computing an ES.</p> <p>Finally, Graham et al. ([<reflink idref="bib64" id="ref83">64</reflink>]) created a description of the applied writing treatments in each study. To categorize writing treatments, they drew on instructional categorizations used in previous meta-analyses (Graham & Perin, [<reflink idref="bib67" id="ref84">67</reflink>]; Graham et al., [<reflink idref="bib66" id="ref85">66</reflink>], [<reflink idref="bib63" id="ref86">63</reflink>]; Hillocks, [<reflink idref="bib78" id="ref87">78</reflink>]; Koster et al., [<reflink idref="bib98" id="ref88">98</reflink>]). Their review included 23 different writing treatments with enough independent treatment/control comparisons that an average weighted ES could be calculated using robust variance estimation (RVE). Because we eliminated 64% of the independent treatment comparisons from Graham et al. ([<reflink idref="bib64" id="ref89">64</reflink>]) in the present review, the number of writing treatments with enough comparisons to create an average weighted ES with RVE was reduced to 10. These treatments are described in Table 1, which also provides a description of which aspect of the WWC model (Graham, [<reflink idref="bib57" id="ref90">57</reflink>]) the writing treatment was likely to enhance.</p> <p>With two exceptions, two trained members of the Graham et al. ([<reflink idref="bib64" id="ref91">64</reflink>]) research team coded each study for all of the variables above. The exceptions involved type of writing or reading measure and the categorization of writing treatments. These exceptions were coded by Graham, who has over 40 years of experience as a writing researcher, and another member of the research team. The percentage of agreement was high for coded variables (94%), and all differences were resolved through discussion.</p> <hd id="AN0184705652-10">Calculation of ESs</hd> <p>When calculating ESs for writing outcome in a study, Graham et al. ([<reflink idref="bib64" id="ref92">64</reflink>]) computed the difference between the average performance of treatment and control students at posttest and dividing this by the pooled standard deviation for the two groups. Before calculating each ES, posttest scores were adjusted by subtracting pretest mean scores from them for each condition. The only exceptions to this pretest adjustment involved studies where randomization occurred but no pretest means were provided. All ESs were calculated as Hedge's <emph>g.</emph> It was necessary in some instances for Graham et al. ([<reflink idref="bib64" id="ref93">64</reflink>]) to reverse the direction of ESs (e.g., the outcome assessed writing errors).</p> <p>For the 1076 ESs included in the present review, we examined them to identify outliers. As specified by Tukey ([<reflink idref="bib157" id="ref94">157</reflink>]), an outlier was defined as any ES falling 1.5 interquartile ranges (75th to 25th percentile) above or below the 25th and 75th percentile. Using this definition, it was necessary to winsorize two ESs of 5.40 and 5.06 in Rijlaarsdam et al. ([<reflink idref="bib137" id="ref95">137</reflink>]), one ES in Lane ([<reflink idref="bib100" id="ref96">100</reflink>]) of 5.41, one in Couzijn and Rijlaarsdam ([<reflink idref="bib25" id="ref97">25</reflink>]) of −1.62, one in Harville ([<reflink idref="bib72" id="ref98">72</reflink>]) of −1.99, and one in Couzijn and Rijlaarsdam ([<reflink idref="bib26" id="ref99">26</reflink>]) of −1.46.</p> <hd id="AN0184705652-11">Statistical analyses</hd> <p>We employed RVE to analyze the ESs included in the present review. A random effects model was specified, and all calculations and statistical analyses involving average weighted ESs as well as meta-regression moderator analyses were conducted with RStudio open-source software (Version 3.6.3; R Core Team, [<reflink idref="bib132" id="ref100">132</reflink>]; Version 1.2.5033; RStudio Team, [<reflink idref="bib143" id="ref101">143</reflink>]) using functions available in the Robumeta package (Version 2.0; Fisher et al., [<reflink idref="bib47" id="ref102">47</reflink>]). Many studies in this review yielded multiple ESs (derived from multiple measures) for the same subsample and had great variability in sample sizes. Therefore, RVE was employed as it provided a method for handling our correlated and dependent data structure with small sample correction. It must be noted that RVE does not model heterogeneity at multiple levels, nor does it provide corresponding hypothesis tests. Additionally, the power of the categorical moderator depends greatly on the number of studies and features of the covariate (Tanner-Smith et al., [<reflink idref="bib151" id="ref103">151</reflink>]). When the number of treatment/control comparisons is small, the test statistics and confidence intervals based on RVE can result in inflated Type I errors (Hedges et al., [<reflink idref="bib73" id="ref104">73</reflink>]; Tipton & Pustejovsky, [<reflink idref="bib155" id="ref105">155</reflink>]). As a result, some of our moderation analysis results must be interpreted with caution as RVE requires a sufficiently large sample size and may yield inaccurate <emph>p</emph> values when degrees of freedom are under 4.</p> <p>For all average weighted ESs, a confidence interval was calculated. <emph>I</emph><sups>2</sups> was calculated to describe the ratio of true heterogeneity to total variance (see Table 2). An <emph>I</emph><sups>2</sups> of 0 to 40% might not be important, 30 to 60% may represent moderate heterogeneity, 50 to 90% can indicate substantial heterogeneity, and 75 to 100% suggests considerable heterogeneity (Higgins & Green, [<reflink idref="bib81" id="ref106">81</reflink>]).</p> <p>Table 2 Average-weighted effect sizes overall and specific writing treatments</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Overall</p></th><th align="left"><p><italic>N/n</italic></p></th><th align="left"><p><italic>k</italic></p></th><th align="left"><p><math xmlns="http://www.w3.org/1998/Math/MathML"><mover xmlns=""><mi>g</mi><mo>¯</mo></mover></math><inline-graphic mime-subtype="GIF" href="11145_2024_10539_Article_IEq1.gif" />(SE)</p></th><th align="left"><p>95% CI</p></th><th align="left"><p><italic>p</italic></p></th><th align="left"><p><italic>I</italic><sup>2</sup></p></th></tr></thead><tbody><tr><td align="left"><p>Writing achievement</p></td><td align="left"><p>148/128</p></td><td align="left"><p>1076</p></td><td align="left"><p>.38***(.04)</p></td><td align="left"><p>[.30,.47]</p></td><td align="left"><p> <.001</p></td><td align="left"><p>86.83</p></td></tr><tr><td align="left" colspan="7"><p><italic>By specific measures</italic></p></td></tr><tr><td align="left"><p>Writing quality</p></td><td align="left"><p>98/83</p></td><td align="left"><p>213</p></td><td align="left"><p>.41***(.06)</p></td><td align="left"><p>[.30,.52]</p></td><td align="left"><p> <.001</p></td><td align="left"><p>84.07</p></td></tr><tr><td align="left"><p>Ideation</p></td><td align="left"><p>48/47</p></td><td align="left"><p>152</p></td><td align="left"><p>.37***(.08)</p></td><td align="left"><p>[.21,.53]</p></td><td align="left"><p> <.001</p></td><td align="left"><p>85.74</p></td></tr><tr><td align="left"><p>Elements</p></td><td align="left"><p>21/21</p></td><td align="left"><p>202</p></td><td align="left"><p>.47*(.18)</p></td><td align="left"><p>[.09,.84]</p></td><td align="left"><p> <.05</p></td><td align="left"><p>93.89</p></td></tr><tr><td align="left"><p>Organization</p></td><td align="left"><p>26/24</p></td><td align="left"><p>70</p></td><td align="left"><p>.27*(.11)</p></td><td align="left"><p>[.04,.51]</p></td><td align="left"><p> <.05</p></td><td align="left"><p>85.51</p></td></tr><tr><td align="left"><p>Voice</p></td><td align="left"><p>12/11</p></td><td align="left"><p>32</p></td><td align="left"><p>.25*(.11)</p></td><td align="left"><p>[.00,.49]</p></td><td align="left"><p> <.05</p></td><td align="left"><p>68.20</p></td></tr><tr><td align="left"><p>Process</p></td><td align="left"><p>11/11</p></td><td align="left"><p>40</p></td><td align="left"><p>.99**(.22)</p></td><td align="left"><p>[.49, 1.49]</p></td><td align="left"><p> <.01</p></td><td align="left"><p>90.68</p></td></tr><tr><td align="left"><p>Translation</p></td><td align="left"><p>26/24</p></td><td align="left"><p>153</p></td><td align="left"><p>.34*(.13)</p></td><td align="left"><p>[.08,.60]</p></td><td align="left"><p> <.05</p></td><td align="left"><p>89.73</p></td></tr><tr><td align="left"><p>Transcription</p></td><td align="left"><p>14/12</p></td><td align="left"><p>45</p></td><td align="left"><p>.44*(.16)</p></td><td align="left"><p>[.10,.78]</p></td><td align="left"><p> <.05</p></td><td align="left"><p>89.40</p></td></tr><tr><td align="left"><p>Motivational beliefs</p></td><td align="left"><p>32/27</p></td><td align="left"><p>145</p></td><td align="left"><p>.07(.04)</p></td><td align="left"><p>[−.01,.16]</p></td><td align="left"><p>.08</p></td><td align="left"><p>19.65</p></td></tr><tr><td align="left" colspan="7"><p><italic>By writing treatment</italic></p></td></tr><tr><td align="left"><p>Process approach</p></td><td align="left"><p>6/6</p></td><td align="left"><p>56</p></td><td align="left"><p>.75*(.27)</p></td><td align="left"><p>[.05, 1.44]</p></td><td align="left"><p> <.05</p></td><td align="left"><p>90.88</p></td></tr><tr><td align="left" colspan="7"><p><italic>By specific measures</italic></p></td></tr><tr><td align="left"><p>Writing quality</p></td><td align="left"><p>5/5</p></td><td align="left"><p>9</p></td><td align="left"><p>1.03(.40)</p></td><td align="left"><p>[−.08, 2.15]</p></td><td align="left"><p>.062</p></td><td align="left"><p>93.41</p></td></tr><tr><td align="left"><p>Strategy instruction</p></td><td align="left"><p>29/24</p></td><td align="left"><p>137</p></td><td align="left"><p>.59***(.11)</p></td><td align="left"><p>[.38,.81]</p></td><td align="left"><p> <.001</p></td><td align="left"><p>89.44</p></td></tr><tr><td align="left"><p><italic>SRSD</italic></p></td><td align="left"><p>8/8</p></td><td align="left"><p>38</p></td><td align="left"><p>.84**(.22)</p></td><td align="left"><p>[.31, 1.36]</p></td><td align="left"><p> <.01</p></td><td align="left"><p>94.39</p></td></tr><tr><td align="left"><p><italic>Not SRSD</italic></p></td><td align="left"><p>22/17</p></td><td align="left"><p>99</p></td><td align="left"><p>.51***(.10)</p></td><td align="left"><p>[.29,.72]</p></td><td align="left"><p> <.001</p></td><td align="left"><p>84.96</p></td></tr><tr><td align="left" colspan="7"><p><italic>By specific measures</italic></p></td></tr><tr><td align="left"><p>Writing quality</p></td><td align="left"><p>23/18</p></td><td align="left"><p>37</p></td><td align="left"><p>.71***(.14)</p></td><td align="left"><p>[.42, 1.00]</p></td><td align="left"><p> <.001</p></td><td align="left"><p>89.05</p></td></tr><tr><td align="left"><p>Ideation</p></td><td align="left"><p>15/15</p></td><td align="left"><p>31</p></td><td align="left"><p>.55*(.20)</p></td><td align="left"><p>[.14,.97]</p></td><td align="left"><p> <.05</p></td><td align="left"><p>91.44</p></td></tr><tr><td align="left"><p>Digital writing</p></td><td align="left"><p>22/19</p></td><td align="left"><p>101</p></td><td align="left"><p>.18(.10)</p></td><td align="left"><p>[−.02,.38]</p></td><td align="left"><p>.081</p></td><td align="left"><p>76.29</p></td></tr><tr><td align="left"><p><italic>Word processing</italic></p></td><td align="left"><p>12/11</p></td><td align="left"><p>72</p></td><td align="left"><p>.08(.10)</p></td><td align="left"><p>[−.15,.31]</p></td><td align="left"><p>.46</p></td><td align="left"><p>51.67</p></td></tr><tr><td align="left"><p><italic>Word processing plus</italic></p></td><td align="left"><p>9/8</p></td><td align="left"><p>25</p></td><td align="left"><p>.26(.17)</p></td><td align="left"><p>[−.14,.66]</p></td><td align="left"><p>.17</p></td><td align="left"><p>85.69</p></td></tr><tr><td align="left" colspan="7"><p><italic>By specific measures</italic></p></td></tr><tr><td align="left"><p>Writing quality</p></td><td align="left"><p>20/17</p></td><td align="left"><p>32</p></td><td align="left"><p>.15(.11)</p></td><td align="left"><p>[−.07,.38]</p></td><td align="left"><p>.171</p></td><td align="left"><p>70.54</p></td></tr><tr><td align="left"><p>Ideation</p></td><td align="left"><p>6/5</p></td><td align="left"><p>12</p></td><td align="left"><p>.26(.15)</p></td><td align="left"><p>[−.14,.66]</p></td><td align="left"><p>.145</p></td><td align="left"><p>55.38</p></td></tr><tr><td align="left"><p>Transcription instruction</p></td><td align="left"><p>12/10</p></td><td align="left"><p>42</p></td><td align="left"><p>.54*(.18)</p></td><td align="left"><p>[.16,.93]</p></td><td align="left"><p>.010</p></td><td align="left"><p>92.07</p></td></tr><tr><td align="left" colspan="7"><p><italic>By specific measures</italic></p></td></tr><tr><td align="left"><p>Transcription</p></td><td align="left"><p>11/9</p></td><td align="left"><p>29</p></td><td align="left"><p>.52*(.19)</p></td><td align="left"><p>[.10,.95]</p></td><td align="left"><p> <.05</p></td><td align="left"><p>91.81</p></td></tr><tr><td align="left"><p>CAI</p></td><td align="left"><p>7/6</p></td><td align="left"><p>16</p></td><td align="left"><p>.05(.07)</p></td><td align="left"><p>[−.13,.24]</p></td><td align="left"><p>.496</p></td><td align="left"><p>0</p></td></tr><tr><td align="left"><p>Feedback</p></td><td align="left"><p>18/16</p></td><td align="left"><p>259</p></td><td align="left"><p>.30*(.14)</p></td><td align="left"><p>[.01,.59]</p></td><td align="left"><p> <.05</p></td><td align="left"><p>87.62</p></td></tr><tr><td align="left" colspan="7"><p><italic>By specific measures</italic></p></td></tr><tr><td align="left"><p>Writing quality</p></td><td align="left"><p>8/8</p></td><td align="left"><p>23</p></td><td align="left"><p>.24(.15)</p></td><td align="left"><p>[−.10,.59]</p></td><td align="left"><p>.139</p></td><td align="left"><p>82.91</p></td></tr><tr><td align="left"><p>Translation</p></td><td align="left"><p>7/6</p></td><td align="left"><p>100</p></td><td align="left"><p>.45(.33)</p></td><td align="left"><p>[−.35, 1.26]</p></td><td align="left"><p>.22</p></td><td align="left"><p>93.79</p></td></tr><tr><td align="left"><p>Pre-writing</p></td><td align="left"><p>12/11</p></td><td align="left"><p>52</p></td><td align="left"><p>.32*(.10)</p></td><td align="left"><p>[.09,.55]</p></td><td align="left"><p> <.05</p></td><td align="left"><p>81.83</p></td></tr><tr><td align="left" colspan="7"><p><italic>By specific measures</italic></p></td></tr><tr><td align="left"><p>Writing quality<sup>1</sup></p></td><td align="left"><p>5/5</p></td><td align="left"><p>7</p></td><td align="left"><p>.32(.20)</p></td><td align="left"><p>[−.23,.87]</p></td><td align="left"><p>.18</p></td><td align="left"><p>84.58</p></td></tr><tr><td align="left"><p>Ideation</p></td><td align="left"><p>7/7</p></td><td align="left"><p>16</p></td><td align="left"><p>.27(.15)</p></td><td align="left"><p>[−.11, 64]</p></td><td align="left"><p>.13</p></td><td align="left"><p>82.13</p></td></tr><tr><td align="left"><p>Peer assistance</p></td><td align="left"><p>13/12</p></td><td align="left"><p>74</p></td><td align="left"><p>.59**(.18)</p></td><td align="left"><p>[.20,.98]</p></td><td align="left"><p> <.01</p></td><td align="left"><p>92.79</p></td></tr><tr><td align="left" colspan="7"><p><italic>By specific measures</italic></p></td></tr><tr><td align="left"><p>Writing quality</p></td><td align="left"><p>9/8</p></td><td align="left"><p>22</p></td><td align="left"><p>.38**(.11)</p></td><td align="left"><p>[.13,.64]</p></td><td align="left"><p> <.01</p></td><td align="left"><p>82.96</p></td></tr><tr><td align="left"><p>Text structure instruction</p></td><td align="left"><p>7/6</p></td><td align="left"><p>33</p></td><td align="left"><p>.48(.29)</p></td><td align="left"><p>[−.24, 1.19]</p></td><td align="left"><p>.156</p></td><td align="left"><p>95.06</p></td></tr></tbody></table> </ephtml> </p> <p>SRSD refers to Self-Regulated Strategy Development model; CAI refers to computer-assisted instruction; <emph>N</emph> refers to the counts of independent samples nested within <emph>n</emph> which represents the counts of studies; <emph>k</emph> refers to the number of effect sizes. Superscript 1 refers to degrees of freedom under 4, which means results should be taken with caution *<emph>p</emph> <.05; **<emph>p</emph> <.01; ***<emph>p</emph> <.001</p> <hd id="AN0184705652-12">Analysis for specific research questions</hd> <p></p> <ulist> <item> <bold> Research Questions 1. </bold> An average weighted ESs was computed for all writing outcomes across all treatment/control comparisons. Average weighted ESs were also computed for nine writing outcomes (see Table 2). There were not enough treatment/control comparisons assessing writing knowledge, the tenth outcome, to compute an average weighted ES for it.</item> <p></p> <item> <bold> Research Question 2. </bold> We used meta-regression applying RVE with random effects to examine possible moderating effects. Two separate meta-regressions were conducted. The first meta-regression involved the three indicators of study quality, dummy coded as publication type (journal article compared to non-journal article), norm-referenced test (yes or no), and treatment fidelity (treatment fidelity established compared to not established). These three predictors were entered into the model together.</item> <p></p> <item> The second meta-regression involved the following three study characteristics: grade level (centered at grade 6), type of student (typical students [full range as well as average and high achieving students] compared to less capable students [poor writers and students with special needs]), and control condition type (contrast coded to compare business-as-usual to alternative treatment, diluted treatment, no treatment, writing alternative treatment, counter treatment, and reading treatment). Again, these predictors were entered into the model together.</item> <p></p> <item> <bold> Research Questions 3. </bold> An average weighted ESs were computed for all 10 writing treatments described in Table 1. This included average weighted ESs for all writing measures combined and each writing outcome separately when the degrees of freedom for a measure was five or greater. Each of these treatments had been tested in at least four independent treatment/control comparisons. For two writing treatments, digital writing (word processing and word processing plus programs) and strategy instruction (SRSD and non-SRSD strategy instruction), separate average weighted ESs were presented for different forms of each treatment.</item> </ulist> <hd id="AN0184705652-13">Publication bias</hd> <p>Possible publication bias involved examining graphical representations of the funnel plot for all outcomes in all studies. An Egger's test was also conducted with the observed ESs to determine the possibility of publication bias.</p> <hd id="AN0184705652-14">Results</hd> <p></p> <hd id="AN0184705652-15">Characteristics of studies</hd> <p>Information on each of the independent writing treatment/control comparisons is presented in Table 3. This includes reference, grade level of students, type of student, writing treatment, control condition, and aggregated ES and variance for writing performance. Elimination of N of 1 studies, investigations where teacher effects were not controlled, attrition was greater than 20%, and writing measures were not reliable resulted in a 66% reduction of the 406 independent treatment/control comparisons and 32% of the ESs from Graham et al. ([<reflink idref="bib64" id="ref107">64</reflink>]).</p> <p>Table 3 Aggregated effect sizes by primary comparison condition</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Author</p></th><th align="left"><p>Grade</p></th><th align="left"><p>Student type</p></th><th align="left"><p>Treatment vs control</p></th><th align="left"><p>Aggregated effect size</p></th><th align="left"><p>Aggregated variance</p></th></tr></thead><tbody><tr><td align="left"><p>Roscoe et al. (<xref ref-type="bibr" rid="bibr140">2019</xref>)</p></td><td align="left"><p>High school (15–17)</p></td><td align="left" /><td align="left"><p>Strategy vs Alternative</p></td><td align="left"><p>0.1</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left" /><td align="left"><p>High school (15–17)</p></td><td align="left" /><td align="left"><p>Feedback<sup>4</sup> vs Alternative</p></td><td align="left"><p>0.15</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>Braaksma (<xref ref-type="bibr" rid="bibr13">2002</xref>)</p></td><td align="left"><p>10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>−0.05</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left" /><td align="left"><p>10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Other</p></td><td align="left"><p>−0.04</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left" /><td align="left"><p>10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>−0.21</p></td><td char="." align="char"><p>0.08</p></td></tr><tr><td align="left"><p>Brantley (<xref ref-type="bibr" rid="bibr15">1991</xref>)</p></td><td align="left"><p>Middle school</p></td><td align="left"><p>Students with special needs</p></td><td align="left"><p>Feedback<sup>5</sup> vs Diluted</p></td><td align="left"><p>0.22</p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left"><p>Brewer (<xref ref-type="bibr" rid="bibr16">2002</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Goals vs BAU</p></td><td align="left"><p>0.14</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Christensen (<xref ref-type="bibr" rid="bibr18">2004</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Transcription vs Counter</p></td><td align="left"><p>0.75</p></td><td char="." align="char"><p>0.15</p></td></tr><tr><td align="left"><p>Combs (<xref ref-type="bibr" rid="bibr21">1977</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Sentence vs BAU</p></td><td align="left"><p>1.32</p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left"><p>Conklin (<xref ref-type="bibr" rid="bibr22">2007</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs Alternative</p></td><td align="left"><p>0.21</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Couzijn and Rijlaarsdam (<xref ref-type="bibr" rid="bibr26">2005</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Observation vs Other</p></td><td align="left"><p>0.09</p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left" /><td align="left"><p>9</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Observation vs Counter</p></td><td align="left"><p>0.53</p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left" /><td align="left"><p>9</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Feedback<sup>5</sup> vs Other</p></td><td align="left"><p>−1.03</p></td><td char="." align="char"><p>0.14</p></td></tr><tr><td align="left" /><td align="left"><p>9</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Feedback<sup>5</sup> vs Counter</p></td><td align="left"><p>−0.28</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>Covill (<xref ref-type="bibr" rid="bibr27">1996</xref>)</p></td><td align="left"><p>10,11</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Feedback<sup>2</sup> vs Counter</p></td><td align="left"><p>0.14</p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left" /><td align="left"><p>10,11</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Feedback<sup>2</sup> vs Counter</p></td><td align="left"><p>0.14</p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left"><p>Dailey (<xref ref-type="bibr" rid="bibr32">1992</xref>)</p></td><td align="left"><p>9,10</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Peer assistance and support vs Counter</p></td><td align="left"><p>1.38</p></td><td char="." align="char"><p>0.09</p></td></tr><tr><td align="left"><p>De La Paz and Graham (<xref ref-type="bibr" rid="bibr34">2002</xref>)</p></td><td align="left"><p>7,8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy<sup>1</sup> vs BAU</p></td><td align="left"><p>0.78</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>de Ment (<xref ref-type="bibr" rid="bibr37">2008</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>High achieving students</p></td><td align="left"><p>Feedback<sup>5</sup> vs No treatment</p></td><td align="left"><p>0.21</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Eliason (<xref ref-type="bibr" rid="bibr43">1994</xref>)</p></td><td align="left"><p>11</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs BAU</p></td><td align="left"><p>−0.1</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Erickson (<xref ref-type="bibr" rid="bibr44">2009</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Increased writing vs Counter</p></td><td align="left"><p>0.02</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>Barton (<xref ref-type="bibr" rid="bibr8">2018</xref>)</p></td><td align="left"><p>11</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Peer assistance and support vs BAU</p></td><td align="left"><p>0.52</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Espinoza (<xref ref-type="bibr" rid="bibr45">1992</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Diluted</p></td><td align="left"><p>0.18</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left" /><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>0.17</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Frank (<xref ref-type="bibr" rid="bibr48">2008</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Transcription vs BAU</p></td><td align="left"><p>0</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>Frost (<xref ref-type="bibr" rid="bibr50">2008</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Digital writing vs Diluted</p></td><td align="left"><p>−0.04</p></td><td char="." align="char"><p>0.17</p></td></tr><tr><td align="left"><p>Galbraith (<xref ref-type="bibr" rid="bibr51">2014</xref>)</p></td><td align="left"><p>9,10</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Strategy<sup>1</sup> vs BAU</p></td><td align="left"><p>0.11</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Hamilton (<xref ref-type="bibr" rid="bibr69">1960</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Transcription vs Counter</p></td><td align="left"><p>0.01</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Harville (<xref ref-type="bibr" rid="bibr72">2001</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Students with special needs</p></td><td align="left"><p>Digital writing vs BAU</p></td><td align="left"><p>−0.92</p></td><td char="." align="char"><p>0.16</p></td></tr><tr><td align="left"><p>Hickerson (<xref ref-type="bibr" rid="bibr74">1987</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Text structure vs BAU</p></td><td align="left"><p>0.71</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left" /><td align="left"><p>10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Text structure vs BAU</p></td><td align="left"><p>−0.13</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>Higgins (<xref ref-type="bibr" rid="bibr75">2013</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Creativity/critical thinking vs BAU</p></td><td align="left"><p>0.39</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Hill (<xref ref-type="bibr" rid="bibr76">1990</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>High achieving students</p></td><td align="left"><p>Peer assistance and support vs Counter</p></td><td align="left"><p>0.06</p></td><td char="." align="char"><p>0.09</p></td></tr><tr><td align="left"><p>Hillocks (<xref ref-type="bibr" rid="bibr79">1982</xref>)</p></td><td align="left"><p>7,8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Revision vs Diluted</p></td><td align="left"><p>−0.25</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left" /><td align="left"><p>7,8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs Diluted</p></td><td align="left"><p>−0.15</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left" /><td align="left"><p>7,8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs Counter</p></td><td align="left"><p>0.41</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left" /><td align="left"><p>7,8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Other vs Counter</p></td><td align="left"><p>0.18</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Hisgen et al. (<xref ref-type="bibr" rid="bibr80">2020</xref>)</p></td><td align="left"><p>8,9</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Strategy<sup>1</sup> vs Alternative</p></td><td align="left"><p>2.18</p></td><td char="." align="char"><p>0.09</p></td></tr><tr><td align="left"><p>Holley (<xref ref-type="bibr" rid="bibr82">1990</xref>)</p></td><td align="left"><p>12</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Feedback<sup>3</sup> vs Counter</p></td><td align="left"><p>2.39</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Holliway and McCutchen (<xref ref-type="bibr" rid="bibr83">2004</xref>)</p></td><td align="left"><p>5,9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>2</sup> vs Diluted</p></td><td align="left"><p>0.43</p></td><td char="." align="char"><p>0.09</p></td></tr><tr><td align="left" /><td align="left" /><td align="left" /><td align="left"><p>Feedback<sup>2</sup> vs BAU</p></td><td align="left"><p>−0.76</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>Hoogeveen (<xref ref-type="bibr" rid="bibr84">2013</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs Counter</p></td><td align="left"><p>1.14</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Jeroski (<xref ref-type="bibr" rid="bibr90">1982</xref>)</p></td><td align="left"><p>8,9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs Counter</p></td><td align="left"><p>−0.05</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Jones (<xref ref-type="bibr" rid="bibr91">1966</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>High achieving students</p></td><td align="left"><p>Transcription vs Diluted</p></td><td align="left"><p>0.99</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Transcription vs Diluted</p></td><td align="left"><p>0.44</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Kaffar (<xref ref-type="bibr" rid="bibr93">1993</xref>)</p></td><td align="left"><p>9–12</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>CAI vs Counter</p></td><td align="left"><p>−0.06</p></td><td char="." align="char"><p>0.08</p></td></tr><tr><td align="left"><p>Kelley (<xref ref-type="bibr" rid="bibr95">1984</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Text structure vs Reading</p></td><td align="left"><p>1.67</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left" /><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Process writing approach vs Reading</p></td><td align="left"><p>1.85</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Kuscenko (<xref ref-type="bibr" rid="bibr99">2018</xref>)</p></td><td align="left"><p>High school</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Revision vs BAU</p></td><td align="left"><p>0.94</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>Lange et al. (<xref ref-type="bibr" rid="bibr101">2009</xref>)</p></td><td align="left"><p>Secondary level</p><p>(13–15)</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Digital writing vs Diluted</p></td><td align="left"><p>0.5</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left" /><td align="left"><p>Secondary level</p><p>(13–15)</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Digital writing vs No treatment</p></td><td align="left"><p>0.07</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>Laysears-Smith (<xref ref-type="bibr" rid="bibr102">2005</xref>)</p></td><td align="left"><p>10–12</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Increased writing vs No treatment</p></td><td align="left"><p>0.02</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Lott (<xref ref-type="bibr" rid="bibr107">1986</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>0.02</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Lyons (<xref ref-type="bibr" rid="bibr108">2002</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>0.81</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Lane (<xref ref-type="bibr" rid="bibr100">2003</xref>)</p></td><td align="left"><p>6–8</p></td><td align="left"><p>Students with special needs</p></td><td align="left"><p>Strategy vs Alternative</p></td><td align="left"><p>2.15</p></td><td char="." align="char"><p>0.15</p></td></tr><tr><td align="left" /><td align="left"><p>6–8</p></td><td align="left"><p>Students with special needs</p></td><td align="left"><p>Strategy vs No treatment</p></td><td align="left"><p>4.41</p></td><td char="." align="char"><p>0.50</p></td></tr><tr><td align="left"><p>Lytle (<xref ref-type="bibr" rid="bibr109">1987</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>−0.05</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Mayo (<xref ref-type="bibr" rid="bibr111">1976</xref>)</p></td><td align="left"><p>10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs Diluted</p></td><td align="left"><p>−0.09</p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left" /><td align="left"><p>10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs No treatment</p></td><td align="left"><p>−0.35</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>McCarty (<xref ref-type="bibr" rid="bibr112">2016</xref>)</p></td><td align="left"><p>11</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs Counter</p></td><td align="left"><p>0.27</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>McCreight (<xref ref-type="bibr" rid="bibr113">1995</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs Other</p></td><td align="left"><p>−0.18</p></td><td char="." align="char"><p>0.09</p></td></tr><tr><td align="left"><p>McDermott (<xref ref-type="bibr" rid="bibr114">2009</xref>)</p></td><td align="left"><p>10–12</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Procedural facilitation vs No treatment</p></td><td align="left"><p>0.23</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>McNeill et al. (<xref ref-type="bibr" rid="bibr115">2006</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Text structure vs Diluted</p></td><td align="left"><p>0.04</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Wilson and Roscoe (<xref ref-type="bibr" rid="bibr169">2019</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Diluted</p></td><td align="left"><p>0.03</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Moseley (<xref ref-type="bibr" rid="bibr117">2003</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Vocabulary vs Diluted</p></td><td align="left"><p>−0.15</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Olina and Sullivan (<xref ref-type="bibr" rid="bibr121">2004</xref>)</p></td><td align="left"><p>10,11</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>25</sup> vs Diluted</p></td><td align="left"><p>0.45</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Pedersen (<xref ref-type="bibr" rid="bibr126">1977</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Sentence vs BAU</p></td><td align="left"><p>0.44</p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left"><p>Pittman (<xref ref-type="bibr" rid="bibr127">2007</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Transcription vs No treatment</p></td><td align="left"><p>1.04</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Pivarnik (<xref ref-type="bibr" rid="bibr128">1985</xref>)</p></td><td align="left"><p>11</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>0.66</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Reynolds and Perin (<xref ref-type="bibr" rid="bibr134">2009</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy<sup>1</sup> vs Diluted</p></td><td align="left"><p>0.84</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Reynolds et al. (<xref ref-type="bibr" rid="bibr133">1988</xref>)</p></td><td align="left"><p>6–8</p></td><td align="left"><p>Students with special needs</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.16</p></td><td char="." align="char"><p>0.54</p></td></tr><tr><td align="left"><p>Rijlaarsdam and Schoonen (<xref ref-type="bibr" rid="bibr136">1988</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>3</sup> vs BAU</p></td><td align="left"><p>−0.09</p></td><td char="." align="char"><p>0.17</p></td></tr><tr><td align="left"><p>Rolfe (<xref ref-type="bibr" rid="bibr139">1991</xref>)</p></td><td align="left"><p>7,8</p></td><td align="left"><p>Students with special needs</p></td><td align="left"><p>Transcription vs Diluted</p></td><td align="left"><p>0.28</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>Roscoe et al. (<xref ref-type="bibr" rid="bibr141">2013</xref>)</p></td><td align="left"><p>High school</p></td><td align="left" /><td align="left"><p>CAI vs Diluted</p></td><td align="left"><p>0.21</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Rosenbluth and Reed (<xref ref-type="bibr" rid="bibr142">1992</xref>)</p></td><td align="left"><p>11</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>−0.1</p></td><td char="." align="char"><p>0.1</p></td></tr><tr><td align="left" /><td align="left"><p>11</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>0.58</p></td><td char="." align="char"><p>0.14</p></td></tr><tr><td align="left"><p>Hoogeveen and van Gelderen (<xref ref-type="bibr" rid="bibr85">2015</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Text structure vs BAU</p></td><td align="left"><p>0.57</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>Sloan (<xref ref-type="bibr" rid="bibr148">2017</xref>)</p></td><td align="left"><p>12</p></td><td align="left"><p>High achieving students</p></td><td align="left"><p>Feedback<sup>3</sup> vs BAU</p></td><td align="left"><p>−0.04</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>Spilton (<xref ref-type="bibr" rid="bibr149">1986</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Sentence vs BAU</p></td><td align="left"><p>0.37</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Other vs Counter</p></td><td align="left"><p>−0.34</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Tezler (<xref ref-type="bibr" rid="bibr152">1993</xref>)</p></td><td align="left"><p>6,7</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Transcription vs Diluted</p></td><td align="left"><p>0.14</p></td><td char="." align="char"><p>0.19</p></td></tr><tr><td align="left" /><td align="left"><p>6,7</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Transcription vs No treatment</p></td><td align="left"><p>1.06</p></td><td char="." align="char"><p>0.22</p></td></tr><tr><td align="left" /><td align="left"><p>6,7</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Transcription vs Counter</p></td><td align="left"><p>0.45</p></td><td char="." align="char"><p>0.19</p></td></tr><tr><td align="left"><p>Thibodeau (<xref ref-type="bibr" rid="bibr153">1964</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Grammar vs BAU</p></td><td align="left"><p>0.33</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left" /><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs Counter</p></td><td align="left"><p>−0.12</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left" /><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Grammar vs BAU</p></td><td align="left"><p>0.36</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left" /><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs Counter</p></td><td align="left"><p>0.11</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>Thomas (<xref ref-type="bibr" rid="bibr154">1995</xref>)</p></td><td align="left"><p>10</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Text structure vs Reading</p></td><td align="left"><p>1.1</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>van Beuningen et al. (<xref ref-type="bibr" rid="bibr159">2012</xref>)</p></td><td align="left"><p>Secondary level</p><p>(14–15)</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>2</sup> vs Diluted</p></td><td align="left"><p>0.01</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left" /><td align="left"><p>Secondary level</p><p>(14–15)</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>2</sup> vs No treatment</p></td><td align="left"><p>0.31</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left" /><td align="left"><p>Secondary level</p><p>(14–15)</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>2</sup> vs Other</p></td><td align="left"><p>0.35</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left" /><td align="left"><p>Secondary level</p><p>(14–15)</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>2</sup> vs Counter</p></td><td align="left"><p>0.39</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left" /><td align="left"><p>Secondary level</p><p>(14–15)</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>2</sup> vs Diluted</p></td><td align="left"><p>−0.03</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left" /><td align="left"><p>Secondary level</p><p>(14–15)</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>2</sup> vs No treatment</p></td><td align="left"><p>0.15</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left" /><td align="left"><p>Secondary level</p><p>(14–15)</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>2</sup> vs Other</p></td><td align="left"><p>0.49</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>van Drie et al. (<xref ref-type="bibr" rid="bibr160">2005</xref>)</p></td><td align="left"><p>High school</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Procedural facilitation vs Counter</p></td><td align="left"><p>−0.71</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>van Wagenen (<xref ref-type="bibr" rid="bibr163">1988</xref>)</p></td><td align="left"><p>11</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>0.06</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>Wagner (<xref ref-type="bibr" rid="bibr165">1978</xref>)</p></td><td align="left"><p>9,10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs No treatment</p></td><td align="left"><p>1.71</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left" /><td align="left"><p>9,10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Transcription vs No treatment</p></td><td align="left"><p>1.86</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left" /><td align="left"><p>9,10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Transcription vs No treatment</p></td><td align="left"><p>0.16</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Wissinger and de La Paz (<xref ref-type="bibr" rid="bibr171">2016</xref>)</p></td><td align="left"><p>6,7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs Counter</p></td><td align="left"><p>0.59</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Wong et al. (<xref ref-type="bibr" rid="bibr173">2008</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy<sup>1</sup> vs Diluted</p></td><td align="left"><p>0.88</p></td><td char="." align="char"><p>0.08</p></td></tr><tr><td align="left"><p>Yeh (<xref ref-type="bibr" rid="bibr174">1998</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs Counter</p></td><td align="left"><p>0.6</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>Zellermayer et al. (<xref ref-type="bibr" rid="bibr175">1991</xref>)</p></td><td align="left"><p>9–11</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Diluted</p></td><td align="left"><p>1.14</p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left"><p>Cremin et al. (<xref ref-type="bibr" rid="bibr28">2020</xref>)</p></td><td align="left"><p>Secondary level</p></td><td align="left" /><td align="left"><p>Creativity/critical thinking vs BAU</p></td><td align="left"><p>−0.02</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Jones et al. (<xref ref-type="bibr" rid="bibr92">2013</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Grammar vs BAU</p></td><td align="left"><p>0.22</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Couzijn and Rijlaarsdam (<xref ref-type="bibr" rid="bibr25">2004</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Observation vs Reading</p></td><td align="left"><p>0.69</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left" /><td align="left"><p>9</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Emulation vs Reading</p></td><td align="left"><p>−0.56</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>Wilson and Czik (<xref ref-type="bibr" rid="bibr168">2016</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Students with special needs</p></td><td align="left"><p>Feedback<sup>4</sup> vs Diluted</p></td><td align="left"><p>0.02</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Segers and Verhoeven (<xref ref-type="bibr" rid="bibr146">2009</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs Counter</p></td><td align="left"><p>0.21</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Braaksma (<xref ref-type="bibr" rid="bibr14">2018</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Observation vs Counter</p></td><td align="left"><p>0.05</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Hoogeveen and Gelderen (<xref ref-type="bibr" rid="bibr86">2018</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs Counter</p></td><td align="left"><p>1.43</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Lee and Schallert (<xref ref-type="bibr" rid="bibr103">2016</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Increased writing vs Reading</p></td><td align="left"><p>0.12</p></td><td char="." align="char"><p>0.08</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Increased writing vs BAU</p></td><td align="left"><p>−0.14</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Increased writing vs Reading</p></td><td align="left"><p>0.03</p></td><td char="." align="char"><p>0.08</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Increased writing vs BAU</p></td><td align="left"><p>0.45</p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Increased writing vs Reading</p></td><td align="left"><p>−0.08</p></td><td char="." align="char"><p>0.08</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Increased writing vs BAU</p></td><td align="left"><p>0.02</p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>High achieving students</p></td><td align="left"><p>Increased writing vs Reading</p></td><td align="left"><p>0.34</p></td><td char="." align="char"><p>0.08</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>High achieving students</p></td><td align="left"><p>Increased writing vs BAU</p></td><td align="left"><p>0.86</p></td><td char="." align="char"><p>0.14</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>High achieving students</p></td><td align="left"><p>Increased writing vs Reading</p></td><td align="left"><p>0.26</p></td><td char="." align="char"><p>0.09</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>High achieving students</p></td><td align="left"><p>Increased writing vs BAU</p></td><td align="left"><p>0.43</p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left"><p>Wise (<xref ref-type="bibr" rid="bibr170">1992</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Revision vs No treatment</p></td><td align="left"><p>0.55</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Revision vs Other</p></td><td align="left"><p>−0.13</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>3</sup> vs No treatment</p></td><td align="left"><p>0.69</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Benson (<xref ref-type="bibr" rid="bibr9">1979</xref>)</p></td><td align="left"><p>7–9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Feedback<sup>3</sup> vs BAU</p></td><td align="left"><p>0.08</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Crook (<xref ref-type="bibr" rid="bibr29">1985</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>0.4</p></td><td char="." align="char"><p>0.55</p></td></tr><tr><td align="left"><p>Crossley et al. (<xref ref-type="bibr" rid="bibr30">2013</xref>)</p></td><td align="left"><p>9–12</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>CAI vs Diluted</p></td><td align="left"><p>−0.24</p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left" /><td align="left"><p>9–12</p></td><td align="left"><p>High achieving students</p></td><td align="left"><p>CAI vs Diluted</p></td><td align="left"><p>0.18</p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left"><p>Crossley et al. (<xref ref-type="bibr" rid="bibr31">2013</xref>)</p></td><td align="left"><p>9–12</p></td><td align="left" /><td align="left"><p>CAI vs Diluted</p></td><td align="left"><p>0.01</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Daiute and Kruidenier (<xref ref-type="bibr" rid="bibr33">1985</xref>)</p></td><td align="left"><p>7–9</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Digital writing vs Diluted</p></td><td align="left"><p>0.89</p></td><td char="." align="char"><p>0.64</p></td></tr><tr><td align="left"><p>Grejda (<xref ref-type="bibr" rid="bibr55">1988</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Counter</p></td><td align="left"><p>0.59</p></td><td char="." align="char"><p>0.56</p></td></tr><tr><td align="left"><p>Hammar (<xref ref-type="bibr" rid="bibr70">1986</xref>)</p></td><td align="left"><p>High school</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>Digital writing vs BAU</p></td><td align="left"><p>−0.31</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>Combs (<xref ref-type="bibr" rid="bibr20">1976</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Sentence vs BAU</p></td><td align="left"><p>0.81</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>Iordanou and Constantinou (<xref ref-type="bibr" rid="bibr88">2015</xref>)</p></td><td align="left"><p>11</p></td><td align="left" /><td align="left"><p>Prewriting vs Counter</p></td><td align="left"><p>1.42</p></td><td char="." align="char"><p>0.2</p></td></tr><tr><td align="left"><p>Jacoby (<xref ref-type="bibr" rid="bibr89">1990</xref>)</p></td><td align="left"><p>6,7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>CAI vs BAU</p></td><td align="left"><p>0.09</p></td><td char="." align="char"><p>0.56</p></td></tr><tr><td align="left"><p>Kennedy (<xref ref-type="bibr" rid="bibr96">2008</xref>)</p></td><td align="left"><p>6–12</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Sentence vs No treatment</p></td><td align="left"><p>0.05</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>Kim et al. (<xref ref-type="bibr" rid="bibr97">2011</xref>)</p></td><td align="left"><p>Secondary level</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.19</p></td><td char="." align="char"><p>0.2</p></td></tr><tr><td align="left"><p>Matuchniak et al. (<xref ref-type="bibr" rid="bibr110">2014</xref>)</p></td><td align="left"><p>Secondary level</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.13</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Niemi et al. (<xref ref-type="bibr" rid="bibr119">2007</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Text structure vs Reading</p></td><td align="left"><p>−0.89</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left" /><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Text structure vs BAU</p></td><td align="left"><p>−0.25</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Olson et al. (<xref ref-type="bibr" rid="bibr122">2012</xref>)</p></td><td align="left"><p>6–12</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs Reading</p></td><td align="left"><p>0.3</p></td><td char="." align="char"><p>0.22</p></td></tr><tr><td align="left"><p>Palumbo and Prater (<xref ref-type="bibr" rid="bibr125">1992</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Weaker students</p></td><td align="left"><p>CAI vs Counter</p></td><td align="left"><p>0.28</p></td><td char="." align="char"><p>0.31</p></td></tr><tr><td align="left"><p>Prata et al. (<xref ref-type="bibr" rid="bibr129">2019</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs Counter</p></td><td align="left"><p>0.76</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>De La Paz and Wissinger (<xref ref-type="bibr" rid="bibr35">2017</xref>)</p></td><td align="left"><p>6,7</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Strategy vs Alternative</p></td><td align="left"><p>0.6</p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left"><p>Vinson (<xref ref-type="bibr" rid="bibr164">1971</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Prewriting vs Diluted</p></td><td align="left"><p>0.57</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left" /><td align="left"><p>9</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Emulation vs Diluted</p></td><td align="left"><p>0.18</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Walker (<xref ref-type="bibr" rid="bibr166">1970</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Transcription vs BAU</p></td><td align="left"><p>−0.01</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>Widvey (<xref ref-type="bibr" rid="bibr167">1971</xref>)</p></td><td align="left"><p>11</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Inquiry vs BAU</p></td><td align="left"><p>0.72</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Chung et al. (<xref ref-type="bibr" rid="bibr19">2021</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Revision vs Diluted</p></td><td align="left"><p>0.16</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Lopez et al. (<xref ref-type="bibr" rid="bibr106">2021</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Strategy vs Diluted</p></td><td align="left"><p>0.81</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left" /><td align="left"><p>6</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Observation vs Diluted</p></td><td align="left"><p>0.72</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Wissinger et al. (<xref ref-type="bibr" rid="bibr172">2021</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.73</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>van Drie et al. (<xref ref-type="bibr" rid="bibr161">2021</xref>)</p></td><td align="left"><p>11</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Process writing approach vs BAU</p></td><td align="left"><p>1.15</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>de Smedt and van Keer (<xref ref-type="bibr" rid="bibr38">2018</xref>)</p></td><td align="left"><p>5,6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs Diluted</p></td><td align="left"><p>0.64</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left" /><td align="left"><p>5,6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>2.95</p></td><td char="." align="char"><p>0.17</p></td></tr><tr><td align="left" /><td align="left"><p>5,6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs Diluted</p></td><td align="left"><p>0.35</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>de Smedt et al. (<xref ref-type="bibr" rid="bibr39">2019</xref>)</p></td><td align="left"><p>5,6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs Diluted</p></td><td align="left"><p>−0.27</p></td><td char="." align="char"><p>0.1</p></td></tr><tr><td align="left" /><td align="left"><p>5,6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.14</p></td><td char="." align="char"><p>0.08</p></td></tr><tr><td align="left" /><td align="left"><p>5,6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs Diluted</p></td><td align="left"><p>0.08</p></td><td char="." align="char"><p>0.1</p></td></tr><tr><td align="left" /><td align="left"><p>5,6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs BAU</p></td><td align="left"><p>0.5</p></td><td char="." align="char"><p>0.1</p></td></tr><tr><td align="left" /><td align="left"><p>5,6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Increased writing vs BAU</p></td><td align="left"><p>0.27</p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left"><p>de Smedt et al. (<xref ref-type="bibr" rid="bibr40">2020</xref>)</p></td><td align="left"><p>5,6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left" /><td align="left"><p>5,6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Peer assistance and support vs Diluted</p></td><td align="left"><p>0.22</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Limpo and Alves (<xref ref-type="bibr" rid="bibr104">2014</xref>)</p></td><td align="left"><p>5,6</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Strategy<sup>1</sup> vs BAU</p></td><td align="left"><p>0.87</p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left"><p>Al Saheb (n.d)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy<sup>1</sup> vs BAU</p></td><td align="left"><p>1.09</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>van Driel et al. (<xref ref-type="bibr" rid="bibr162">2022</xref>)</p></td><td align="left"><p>10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs Counter</p></td><td align="left"><p>0.53</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>Page-Voth and Graham (<xref ref-type="bibr" rid="bibr124">1999</xref>)</p></td><td align="left"><p>8,9</p></td><td align="left"><p>Students with special needs</p></td><td align="left"><p>Goals vs Diluted</p></td><td align="left"><p>0.23</p></td><td char="." align="char"><p>0.19</p></td></tr><tr><td align="left" /><td align="left"><p>8,9</p></td><td align="left"><p>Students with special needs</p></td><td align="left"><p>Goals vs BAU</p></td><td align="left"><p>0.84</p></td><td char="." align="char"><p>0.22</p></td></tr><tr><td align="left"><p>Corey (<xref ref-type="bibr" rid="bibr23">1990</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Other</p></td><td align="left"><p>0.62</p></td><td char="." align="char"><p>0.09</p></td></tr><tr><td align="left" /><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Digital writing vs Other</p></td><td align="left"><p>0.07</p></td><td char="." align="char"><p>0.09</p></td></tr><tr><td align="left"><p>Festas et al. (<xref ref-type="bibr" rid="bibr46">2015</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy<sup>1</sup> vs BAU</p></td><td align="left"><p>0.23</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Fraznke et al. (<xref ref-type="bibr" rid="bibr49">2005</xref>)</p></td><td align="left"><p>8</p></td><td align="left"><p>Combinations</p></td><td align="left"><p>Feedback<sup>4</sup> vs Counter</p></td><td align="left"><p>0.22</p></td><td char="." align="char"><p>0.04</p></td></tr><tr><td align="left"><p>Ganong (<xref ref-type="bibr" rid="bibr52">1974</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Process writing approach vs Other</p></td><td align="left"><p>0.1</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>González-Lamas et al. (<xref ref-type="bibr" rid="bibr54">2016</xref>)</p></td><td align="left"><p>Secondary level</p><p>(14–16)</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs Diluted</p></td><td align="left"><p>0.08</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Olson et al. (<xref ref-type="bibr" rid="bibr123">2017</xref>)</p></td><td align="left"><p>7</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.28</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left" /><td align="left"><p>8</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.54</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left" /><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.55</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left" /><td align="left"><p>10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.88</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left" /><td align="left"><p>11</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.43</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left" /><td align="left"><p>12</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Strategy vs BAU</p></td><td align="left"><p>0.05</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Rapanta (<xref ref-type="bibr" rid="bibr131">2021</xref>)</p></td><td align="left"><p>7,8,10</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Prewriting vs BAU</p></td><td align="left"><p>0.54</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Kasparek (<xref ref-type="bibr" rid="bibr94">1994</xref>)</p></td><td align="left"><p>9–11</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Increased writing vs Alternative</p></td><td align="left"><p>0.21</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Rice (<xref ref-type="bibr" rid="bibr135">1968</xref>)</p></td><td align="left"><p>6</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Transcription vs BAU</p></td><td align="left"><p>0.21</p></td><td char="." align="char"><p>0.44</p></td></tr><tr><td align="left"><p>Vahidi et al. (<xref ref-type="bibr" rid="bibr158">2016</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Grammar vs BAU</p></td><td align="left"><p>1.24</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>De La Paz et al. (in press)</p></td><td align="left"><p>10–12</p></td><td align="left" /><td align="left"><p>Strategy vs Counter</p></td><td align="left"><p>0.25</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Adams (<xref ref-type="bibr" rid="bibr1">1971</xref>)</p></td><td align="left"><p>12</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Process writing approach vs BAU</p></td><td align="left"><p>0.3</p></td><td char="." align="char"><p>0.05</p></td></tr><tr><td align="left"><p>Rijlaarsdam et al. (<xref ref-type="bibr" rid="bibr137">2006</xref>)</p></td><td align="left"><p>9</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Feedback<sup>35</sup> vs Other</p></td><td align="left"><p>0.27</p></td><td char="." align="char"><p>0.15</p></td></tr><tr><td align="left" /><td align="left"><p>9</p></td><td align="left"><p>Average students</p></td><td align="left"><p>Procedural facilitation vs Other</p></td><td align="left"><p>0.44</p></td><td char="." align="char"><p>0.14</p></td></tr><tr><td align="left"><p>Barrot (<xref ref-type="bibr" rid="bibr7">2018</xref>)</p></td><td align="left"><p>Secondary level</p><p>(16–18)</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Process writing approach vs BAU</p></td><td align="left"><p>0.81</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>Berman (<xref ref-type="bibr" rid="bibr10">1994</xref>)</p></td><td align="left"><p>11–12</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Process writing approach vs No treatment</p></td><td align="left"><p>0.32</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Black (<xref ref-type="bibr" rid="bibr11">1995</xref>)</p></td><td align="left"><p>10,11</p></td><td align="left"><p>Full classroom</p></td><td align="left"><p>Other vs Counter</p></td><td align="left"><p>0.06</p></td><td char="." align="char"><p>0.04</p></td></tr></tbody></table> </ephtml> </p> <p>Participants' ages are shown in parentheses for the studies that did not report grade levels Full classroom refers to students in regular class. Combinations refer to having students of multiple categories Digital writing includes word processing, word processing plus, and speech to text synthesis. Transcription includes typing and spelling. CAI refers to computer-assisted instruction. Goals include setting goals and goals plus feedback Control types are: BAU = Business as usual, Alternative = a non-writing or non-reading intervention provided, Diluted = the control condition receives part of the treatment condition; this would include where two treatments are identical except one has something the other does not, No treatment = no instruction provided or researchers do not indicate any instruction provided, Counter = there is a counter condition where two treatments are the same, but one has a different mode or delivery of instruction such as word processing vs handwriting; peer vs individual learning; the same material is taught but delivered/taught in a different way such as individual vs group instruction, computer-assisted vs traditional, Reading = the control condition focused on reading or teaching reading, and Other = other types of control conditions <sups>1</sups>Strategy instruction involving SRSD <sups>2</sups>Feedback from teacher <sups>3</sups>Feedback from peer <sups>4</sups>Feedback from computer/machine <sups>5</sups>Feedback from self</p> <p>The earliest study from the 128 documents included in this review was a thesis conducted in 1960 by Hamilton examining the effects of spelling instruction. During the 1960s and 1970s, we located only 4 and 11 studies, respectively, that met our inclusion/exclusion criteria. During the 1980s and 1990s, there was a sizeable increase in writing intervention studies with secondary students (15 and 23 documents, respectively), but such research became even more common in subsequent decades: 30 documents in the 2000s, and 34 in the 2010s.</p> <p>In all, 69% (<emph>n</emph> = 88) of the documents reported on studies conducted in U.S. The most common locations outside of the U.S. were in order: Europe (<emph>n</emph> = 31), Middle East (<emph>n</emph> = 3), Asia, Australia, and Canada (<emph>n</emph> = 2 for each). In 27 of the documents (21%), the teaching of writing involved a language other than English. All but one involved an alphabetic language (one document included a study on teaching writing in Chinese). Although the type of community where studies were conducted was not reported in 37 documents (29%), when this was reported urban communities were the most common (46%), followed by suburban (36%), rural (4%), and some combination of communities (13%). Researchers only indicated SES in 16% of documents. In these cases, low and middle SES communities were most common (43% and 38%, respectively), with the remaining studies mostly occurring in a mix of SES communities (19%).</p> <p>The studies contained in the 128 documents involved 22,838 students in Grades 6 to 12. Grade level (or age) of students was provided in all documents. Forty-five percent of studies involved middle school students, 41% high school students, and 14% both. Most studies were conducted with students in a single grade (59%). Seventeen percent of studies involved two grades, 9% involved three grades, and 14% involved four or more grades. Most documents (96%) also provided enough information to classify participating students. When this was the case, student samples mostly reflected the full range of students in typical classrooms (67%), with the remaining samples split between average writers (7%), high achieving writers (3%), less capable writers (5%), students with special needs (5%), and some combination of these categories (14%).</p> <p>Student gender, race, and SES were not reported in 34%, 48%, and 83% of documents, respectively. When gender was provided, 51% of students were girls. When race was reported, 49% of students were White, 28% Hispanic, 11% Asian, 10% Black, less than 1% First People, and less than 2% Other (e.g., mixed-race). Only 37% of the documents indicated that participating students were L2 or both L2 and L1. When this occurred, 17% of the documents focused just on L2 students.</p> <p>The writing treatments tested in the obtained documents were varied. Across all writing interventions, the most common treatments were strategy instruction (17%), feedback (13%), digital writing (13%), peer assistance and support (7%), transcription instruction (7%), increased writing (7%), pre-writing activities (7%), text structure instruction (4%), computer-assisted instruction (4%), process writing approach (3%), observation (3%), sentence instruction (3%), revision (3%), grammar instruction (2%), goal setting (2%),procedural facilitation (2%), and other writing treatments. These other writing treatments included procedures such emulation of models, inquiry, vocabulary, and critical thinking. The most common control condition was business-as-usual (BAU, 31%), counter (24%), and diluted writing treatment (20%), with other control conditions occurring less often: no treatment (9%), writing alternative treatment (6%), reading treatment (6%), and alternative writing treatment (4%).</p> <p>More than one-half of the documents (59%) provided enough information to determine the amount of time spent teaching writing. When this occurred, an average of 11.74 h was devoted to instruction (<emph>SD</emph> = 8.67 h).</p> <p>It was possible in 99% of documents to determine the content area in which a targeted writing treatment was tested. This mostly occurred in language arts (82%), followed by social studies (9%), science (7%), math (< 2%), and other content areas (< 1%). Likewise, it was possible to determine who led instruction in 78% of documents. These mostly involved teachers (38%), followed by research staff (17%); teachers and technology (12%); peers (9%); teachers and research staff (7%); teacher, research staff, and technology (7%); technology (6%); and research staff and technology (4%).</p> <p>It was further possible in 75% of documents to determine if students engaged in writing during the writing treatment and, if they did so, in what genre. When this occurred, students in the treatment wrote as part of instruction in 87% of the documents. The most common single forms of writing they were asked to undertake were persuasive (40%) and expository (16%), followed by the singular focus on other forms of writing such as summary writing, journal entries, poems (9%), and narrative writing (7%). Slightly more than one-fourth of documents (28%) involved writing in multiple genres. Also, professional development or training was provided to instructors in 73% of the 128 documents.</p> <p>The most common measures in studies were writing quality (20%), genre elements (19%), translation (14%), ideation (14%), writing motivation (14%), organization (7%), transcription (4%), writing process (4%), voice (3%), and writing knowledge (2%).</p> <hd id="AN0184705652-16">Quality of studies</hd> <p>Forty-three percent of studies were published in a research journal. Of the remaining studies, 52% were dissertations and thesis, 2% were book chapters, 2% were unpublished studies from authors, and 2% were research reports and ERIC documents. Seventy percent of studies were experiments, whereas 30% were quasi-experiments with pretests.</p> <p>Treatment fidelity was only established in 38% of studies. When it was established, this mostly occurred through a combination of observation, teacher, and researcher reports (44%) followed by observation (35%), researcher claims (15%), and teacher reports (6%).</p> <hd id="AN0184705652-17">Research question 1: does teaching writing improve writing outcomes?</hd> <p>As can be seen in Table 3, when all writing outcomes were considered collectively, teaching writing resulted in a statistically significant improvement in students' performance (ES = 0.38, and 72% of ESs were positive). Further, teaching writing resulted in statistically detectable improvements for all but one of the nine writing outcomes: writing processes (ES = 0.99), genre elements (0.47), transcription (0.44), writing quality (0.41), ideation (0.37), translation (0.34), organization (0.27), and voice (0.25). Teaching writing did not lead to significant improvement in writing motivation (ES = 0.07).</p> <hd id="AN0184705652-18">Research question 2: do study attributes predict the magnitude of writing outcomes?</hd> <p>For all writing outcomes collectively, two moderator analyses were conducted to determine if study attributes were related to variability in effects. There was considerable variability in ESs, as 87% of the variance was due to between-study factors (see Table 4).</p> <p>Table 4 Moderation analyses on study quality indicators and study characteristics for all measures</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Writing outcomes</p></th><th align="left"><p><italic>k</italic></p></th><th align="left"><p>Estimate (SE)</p></th><th align="left"><p><italic>I</italic><sup>2</sup></p></th></tr></thead><tbody><tr><td align="left"><p>Full model—study quality indicators</p></td><td align="left"><p>1075</p></td><td align="left" /><td char="." align="char"><p>86.96</p></td></tr><tr><td align="left"><p> Intercept</p></td><td align="left" /><td align="left"><p>.34***(.07)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Journal</p></td><td align="left" /><td align="left"><p>.01(.09)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Norm-referenced test</p></td><td align="left" /><td align="left"><p>−.18(.13)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Treatment fidelity established</p></td><td align="left" /><td align="left"><p>.12(.09)</p></td><td char="." align="char" /></tr><tr><td align="left"><p>Full model—study characteristics</p></td><td align="left"><p>990</p></td><td align="left" /><td char="." align="char"><p>87.07</p></td></tr><tr><td align="left"><p> Intercept</p></td><td align="left" /><td align="left"><p>.85**(.27)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Grade level</p></td><td align="left" /><td align="left"><p>−.05(.03)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Less capable writers</p></td><td align="left" /><td align="left"><p>.07(.16)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Control counter</p></td><td align="left" /><td align="left"><p>−.02(.12)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Control other</p></td><td align="left" /><td align="left"><p>−.23(.12)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Control diluted</p></td><td align="left" /><td align="left"><p>−.15(.11)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Control no treatment</p></td><td align="left" /><td align="left"><p>.19(.22)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Control reading</p></td><td align="left" /><td align="left"><p>.02(.33)</p></td><td char="." align="char" /></tr><tr><td align="left"><p> Control alternative<sup>1</sup></p></td><td align="left" /><td align="left"><p>.06(.19)</p></td><td char="." align="char" /></tr></tbody></table> </ephtml> </p> <p> <sups>1</sups> <emph>df</emph> < 4, results should be taken with caution <sups>*</sups><emph>p</emph> <.05; **<emph>p</emph> <.01; ***<emph>p</emph> <.001</p> <p>The first moderator analysis examined if measures of study quality (publication type, norm-referenced test, treatment fidelity) were related to the magnitude of writing effects (Table 3). None of these predictors were uniquely and statistically associated with variation in writing effects.</p> <p>The second moderator analysis examined if specific study characteristics (grade level, type of student, and type of control condition) were related to the magnitude of writing effects. None of these predictors were uniquely or statistically related to variation in writing effects.</p> <hd id="AN0184705652-19">Research question 3: do specific writing treatments improve writing outcomes?</hd> <p></p> <hd id="AN0184705652-20">Average-weighted ES for 10 specific writing treatments</hd> <p>The outcomes for the 10 different writing treatments are presented in Table 2. We presented the average weighted ES for each writing treatment based on all writing outcomes testing that treatment followed by separate ESs for each specific writing outcome when there were at least five or more treatment/control comparisons.</p> <hd id="AN0184705652-21">Comprehensive writing programs (process approach to writing)</hd> <p>Comprehensive writing programs had a statistically detectable impact on students' writing when all writing outcomes were aggregated (ES = 0.75; 84% of all ESs were positive). It should be noted that only six studies employed comprehensive programs (all using process writing approaches) among which five assessed writing quality specifically yielding an average weighted ES of 1.03, but this effect was not statistically significant. Nevertheless, there was considerable variability in ESs (90.88% for all measures and 93.41% for writing quality).</p> <p>These studies involved typical writers in Grades 6 to 12, mostly at the high school level. Three studies took place in the U.S., one in the Netherlands, one in Iceland, and one in the Philippines. Half of the studies applied an experimental design. Half of the control conditions were BAU, and the rest were reading, other, and no treatment.</p> <hd id="AN0184705652-22">Strategy instruction</hd> <p>Strategy instructional studies focused on teaching students to plan text (Yeh, [<reflink idref="bib174" id="ref108">174</reflink>]), edit and/or revise it (Reynolds, [<reflink idref="bib133" id="ref109">133</reflink>]), or both (De La Paz & Graham, [<reflink idref="bib34" id="ref110">34</reflink>]). Teaching students such strategies had a statistically detectable impact on all writing outcomes collectively (ES = 0.59; 79% of all ESs were positive), writing quality (ES = 0.71; 95% of ESs were positive), and ideation (ES = 0.55; 88% of ESs were positive). The variability of ESs was large (89.44%, 89.05%, and 91.44%, respectively).</p> <p>For all studies involving just SRSD strategy instruction, a statistically detectable ES of 0.84 for all writing measures (92% of ESs were positive) was obtained. Variability in ESs was substantial (94.39%). Most SRSD studies involved typical writers in Grades 6 to 12, but average and below-average students were the focus of instruction in three out of eight studies. Six studies took place outside the U.S.: Germany, Canada, Lebanon, and Portugal. Seven out of eight studies applied an experimental design, and the most common control condition was BAU.</p> <p>The remaining strategy studies (not SRSD) resulted in a statistically detectable ES of 0.51 for all writing outcomes (74% of ESs were positive). Variability in ESs was large (84.96%). These studies mostly involved typical writers in the U.S. in Grades 6 to 12, but close to one-third of the studies took place in Spain and Belgium, with close to one-fourth of studies focused on less capable writers or students with special needs. Nineteen out of 22 studies applied an experimental design. The most control conditions were BAU, diluted, and alternative.</p> <hd id="AN0184705652-23">Digital writing</hd> <p>Collectively, the obtained effects for digital writing studies (word processing, word processing plus, and speech-to-text synthesis) were not statistically greater than no effect for all writing measures (ES = 0.18), writing quality (0.15), or ideation (0.26). There was great variability in effects (76.29%, 7.54%, and 55.38%, respectively). In general, students in the digital writing studies were mostly typical writers in Grades 6 to 12 students in the U.S., but one-third of studies focused on less capable writers, and three studies were from Netherlands, Israel, and England. Sixteen out of 22 studies applied an experimental design. The most common control conditions were diluted and counter treatments.</p> <p>When we examined just word processing studies, a statistically nonsignificant effect was also found (0.08) for all writing measures (variability in ESs was 51.67%). This was also the case for word processing plus studies, which yielded a statistically nonsignificant ES of 0.26 for all writing measures (0.26; variability in ESs was 85.69%). Examples of treatments in the word processing plus program studies included multi-media word processing that allowed writers to compose with text, graphics, sound, and videos (Wilson & Roscoe, [<reflink idref="bib169" id="ref111">169</reflink>]) and word processing programs with added software such as a tool that assisted with homophones in writing (Lange et al., [<reflink idref="bib101" id="ref112">101</reflink>]).</p> <hd id="AN0184705652-24">Teaching transcription skills</hd> <p>Studies assessing the effectiveness of teaching transcription skills mostly involved teaching spelling to students in Grades 6 to 10, but one study focused on keyboarding with students in Grade 8. Transcription instruction yielded a statistically detectable ES of 0.54 for all writing measures. Most ESs were positive (93%), and variability was 92.07%. Most of these studies took place in the U.S., but one occurred in Australia. More than half of the studies involved typical writers and the remaining were less capable writers or students with special needs. Five out of twelve studies applied an experimental design, and the most common control conditions were diluted and no treatment. Students did not compose text in any of these studies.</p> <hd id="AN0184705652-25">Computer-assisted instruction</hd> <p>Examples of computer-assisted instruction include an intelligent computer system that offered students help with planning, drafting, and revising (Crossley et al., [<reflink idref="bib31" id="ref113">31</reflink>]) as well as computer-assisted instruction for basic writing skills (Kaffar, [<reflink idref="bib93" id="ref114">93</reflink>]). Such instruction did not result in a statistically detectable ES for all writing outcomes (0.05). These studies (<emph>N</emph> = 6) were conducted mostly with students in the U.S. who were in Grades 6 to 12 (one study was from Australia). Only three studies reported students' ability information and two involved weaker writers. Half of the studies employed an experimental design, and the most common control conditions were diluted treatments. Students wrote in half of these studies.</p> <hd id="AN0184705652-26">Feedback</hd> <p>The feedback produced a statistically detectable ES of 0.30 for all writing measures (64% of all ESs were positive) but not for writing quality (0.24) or translation skills (0.45). Variability in ESs was 87.62%, 82.91%, and 93.79%, respectively. These studies mostly involved typical writers in Grades 6 to 12 in the U.S., although some studies were conducted in the Netherlands and Latvia. Two-thirds of studies used an experimental design, and control conditions were mostly counter and diluted treatments.</p> <hd id="AN0184705652-27">Pre-writing activities</hd> <p>Examples of pre-writing included brainstorming ideas and small group discussions (Vinson, 1980), graphic organizer to generate and organize ideas (Conklin, [<reflink idref="bib22" id="ref115">22</reflink>]), and reading to obtain writing content (van Driel et al., [<reflink idref="bib162" id="ref116">162</reflink>]). Such activities resulted in a statistically detectable ES of 0.32 for all writing outcomes (65% of ESs were positive) but not for writing quality (0.32) or ideation (0.27). Variability in ESs was 81.83%, 84.58%, and 82.13%, respectively. Studies mostly involved typical writers in Grades 6 to 11 and were conducted in the U.S., but some occurred in Canada, Netherlands, Cyprus, and Portugal. All but one study applied an experimental design, and the most common controls were counter and diluted. Students composed text in all studies.</p> <hd id="AN0184705652-28">Peer assistance</hd> <p>Peer assistance mostly involved students working with another student to plan, edit, and/or revise text (e.g., De Smedt et al., [<reflink idref="bib39" id="ref117">39</reflink>]; Prata et al., [<reflink idref="bib129" id="ref118">129</reflink>]). Peers helping each other produced statistically detectable ESs of 0.59 for all writing outcomes (78% of ESs were positive) and 0.38 for writing quality (77% of ESs were positive). Variability in ESs was 92.79% and 82.96%, respectively. Studies mostly included typical students in Grades 6 to 11, although less capable writers were included in two studies. One-half of the studies occurred in the U.S., with the others conducted in Belgium, the Netherlands, and Portugal. Additionally, one-half of studies applied an experimental design, and the most common control conditions were counter-treatments. Students composed text in all but one of the studies.</p> <hd id="AN0184705652-29">Text structure instruction</hd> <p>Examples of such instruction included teaching basic structural elements of a persuasive text (Hoogeveen & van Gelderen, [<reflink idref="bib85" id="ref119">85</reflink>]) and teaching elements for writing an essay in response to material read (Niemi et al., [<reflink idref="bib119" id="ref120">119</reflink>]). Such instruction did not produce a statistically detectable ES for all writing outcomes (0.48), although variability in ESs was considerable (95.06%). Studies mostly included typical Grades 6 to 10 students in the U.S. (one involved weaker students and one occurred in the Netherlands). All but one study applied an experimental design, and the most common control conditions were BAU and reading treatments. Students composed text in all comparisons.</p> <hd id="AN0184705652-30">Increased writing</hd> <p>Examples of increased writing included writing an additional essay each week (Lee & Schallert, [<reflink idref="bib103" id="ref121">103</reflink>]), and writing descriptive text over 10, 50-min sessions (De Smedt et al., [<reflink idref="bib39" id="ref122">39</reflink>]). Such instruction did not result in ES statistically greater than no effect for all writing outcomes (0.14). The impact of increased writing was mostly tested with typical writers in Grades 6 to 12 in the U.S. (one study included less capable writers; one study from South Korea and one from Belgium). All but one study applied an experimental design, and the most common control conditions were BAU and reading treatments.</p> <hd id="AN0184705652-31">Publication bias</hd> <p>A funnel plot was created for all obtained outcomes (see Fig. 1). The funnel plot appeared to be asymmetrical, with the possibility of missing ESs on the left side of the forest plot. The result from Egger's test was consistent with a possible concern about publication bias because the intercept significantly deviated from zero (−1.28; <emph>p</emph> < 0.001).</p> <p>Graph: Fig. 1 Funnel plot</p> <hd id="AN0184705652-32">Discussion</hd> <p>The present review involved a reanalysis of writing intervention outcomes for students in Grades 6 to 12 from a recent meta-analysis conducted by Graham et al. ([<reflink idref="bib64" id="ref123">64</reflink>]). In the prior review, experimental and quasi-experimental writing intervention studies (with pretests) were meta-analyzed to determine if teaching writing improved students' writing performance, what specific writing treatments were effective, and whether writing outcomes were moderated by study features including study quality indicators. In the present best evidence meta-analysis, we applied more stringent standards for study inclusion than Graham et al. ([<reflink idref="bib64" id="ref124">64</reflink>]), eliminating studies from the original review that did not meet specific study quality criteria (e.g., teacher effects were not controlled).</p> <hd id="AN0184705652-33">Teaching writing is a pathway to better writing</hd> <p>When the ESs from the 1,076 ESs for writing outcomes from the 148 independent treatment control comparisons involving 22,838 students in Grades 6 to 12 were examined collectively, teaching writing resulted in an almost four-tenths of a SD improvement (ES = 0.38) in students' writing across all measures. As predicted, this statistically significant effect was smaller than the statistically detectable effect of 0.49 obtained by Graham et al. ([<reflink idref="bib64" id="ref125">64</reflink>]). In contrast to the current best evidence meta-analysis, the current review eliminated studies that were N of 1, did not control for teacher effects, evidenced attrition greater than 20%, and did not include at least one reliable measure of writing. Writing measures that were not established as reliable were also eliminated from any remaining investigations.</p> <p>Our approach for controlling for possible bias due to the study quality features by eliminating specific studies and outcomes significantly reduced the sample of treatment/control comparisons available in this review. Only 34% of the 406 independent treatment/control comparisons and 32% of the ESs for writing available in Graham et al. ([<reflink idref="bib64" id="ref126">64</reflink>]) were still applicable to the analyses conducted in this review. While teaching writing was still an effective approach for improving middle and high school students' writing when methodologically stronger studies were included in the analysis reported here, the sheer volume of eliminated investigations makes it clear that more high-quality writing intervention research is needed. The relative lack of high-quality writing intervention studies is not due to a single factor, but a variety of issues, including a lack of investment by individual countries to fund such research, the failure of universities to adequately prepare Ph.D. candidates to conduct high quality writing research, and the difficulties of conducting research in social contexts (Graham & Harris, [<reflink idref="bib61" id="ref127">61</reflink>]). If we are to provide teachers, policy makers, and other researchers with useful and reliable findings about how to teach writing effectively, we need to develop systemic solutions where invested stakeholders work together to improve the quality of writing intervention research. Because of the cost and difficulty of conducting high-quality intervention research in school contexts, it is important that research funding agencies, professional organizations focused on writing and literacy, policy makers, scholars, administrators, and teachers be involved in reforms to improve writing intervention research.</p> <p>Teaching writing not only enhanced writing when all outcomes were considered collectively in the present review, it also improved specific aspects of writing. This was similar to Graham et al. ([<reflink idref="bib64" id="ref128">64</reflink>]), but as predicted, ESs for specific writing outcomes were generally smaller in this more selective review than the previously more expansive one. This is illustrated below in the ESs presented in parentheses, with the ES from this review presented first and followed by the ES from Graham et al. ([<reflink idref="bib64" id="ref129">64</reflink>]). In terms of what students wrote, statistically detectable improvements were found in this meta-analysis for overall writing quality (ES = 0.41; 0.46), text organization (ES = 0.27; 0.51), inclusion of genre elements (ES = 0.47; 0.60), and voice (ES = 0.25; 0.30). There were also statistically detectable improvements in the production processes writers draw on when composing as described by the WWC model (Graham, [<reflink idref="bib57" id="ref130">57</reflink>]): ideation (ES = 0.37; 0.46), translation skills (ES = 0.34; 0.51), transcription skills (ES = 0.44; 0.49), and strengthened conceptualization and reconceptualization processes (ES = 0.99; 0.82).</p> <p>Contrary to predictions, teaching writing did not produce statistically detectable improvements in motivational beliefs in the present review (ES = 0.07), but statistically positive effects (ES = 0.13) were obtained in Graham et al. ([<reflink idref="bib64" id="ref131">64</reflink>]). There were not enough studies with writing knowledge in the current meta-analysis for us to calculate an effect. The statistically detectable improvement in writing knowledge in Graham et al. ([<reflink idref="bib64" id="ref132">64</reflink>]) was 0.94.</p> <p>The most obvious explanation for the smaller effects obtained for writing collectively as well as for specific writing outcomes in this review and the former one by Graham et al. ([<reflink idref="bib64" id="ref133">64</reflink>]) is that methodologically stronger studies produce smaller effects than methodologically weaker ones (Slavin & Madden, [<reflink idref="bib147" id="ref134">147</reflink>]). All of the effects for writing outcomes were smaller in this more selective review, with just one exception. Measure assessing the writing production processes of conceptualization and reconceptualization yielded an ES of 0.99 in this review, but this ES was just 0.82 in the prior one. The reason for this unexpected outcome is not clear, but it is worth noting that in both reviews the teaching of writing had the strongest impact on this writing outcome.</p> <p>Theoretically, the findings from the current review provided support for specific aspects of the writing process as constructed in the WWC model (Graham, [<reflink idref="bib57" id="ref135">57</reflink>]). This includes the production processes involved in creating text (conceptualization, ideation, translation, transcription, and reconceptualization), as teaching writing resulted in improvements in each of these processes. The findings also support the presumed importance in the WWC model of other cognitive resources including executive control (improvements in conceptualization and reconceptualization are dependent on the application of control processes such as goal setting, planning, monitoring, and evaluation) and LTM (improvements in transcription and translation skills are dependent on spelling, grammar, and sentence knowledge held in LTM).</p> <hd id="AN0184705652-34">Biasing effects for targeted study features were not evident</hd> <p>In Graham et al. ([<reflink idref="bib64" id="ref136">64</reflink>]), three study quality features evidenced a unique and statistically significant correlation with the writing outcomes from the 406 treatment/control comparisons examined in that review. In the current meta-analysis, we controlled for one of these indicators by eliminating studies that were N of 1. We did not, however, control for the other two quality indicators (publication type and norm-referenced measure) that uniquely moderated magnitude of writing effects in the prior meta-analysis. Studies that were not published in a journal were not eliminated because this would overestimate the effects of teaching writing because reviewers are more likely to recommend for publication studies with higher effects. Assessments that were not norm-referenced were also not eliminated because only 7% of all writing assessments involved norm-referenced assessments.</p> <p>In contrast to Graham et al. ([<reflink idref="bib64" id="ref137">64</reflink>]), we did not find that the study quality indicators of publication type or norm-referenced test accounted for statistically detectable unique variance in writing outcomes in this review. Additionally, establishing treatment fidelity did not statistically and uniquely moderate magnitude of writing effects. As in Graham et al. ([<reflink idref="bib64" id="ref138">64</reflink>]), we also did not find that a unique and statistically significant relation between writing outcomes and the three targeted study characteristics of grade-level, type of students (more and less capable writers), and type of control condition.</p> <p>While our sample of treatment/control comparisons in the current review was much smaller than the sample in Graham et al. ([<reflink idref="bib64" id="ref139">64</reflink>]) and we were unable to examine the effects of as many moderators, our findings suggest that the effects of teaching writing are relatively constant across the studies reviewed, at least for the moderators examined here. Thus, the potential separate biasing effects of each of the six predictors included in our analyses were small.</p> <hd id="AN0184705652-35">Secondary students' writing can be improved in multiple ways</hd> <p>Ten specific writing treatments were tested enough times (i.e., four or more independent treatment/control comparisons) in the current review for us to calculate an average weighted ES for them. When we focused on all writing outcomes collectively, six of these treatments yielded statistically detectable improvements in writing: process approach to writing (ES = 0.75), writing strategy instruction (ES = 0.59), transcription instruction (ES = 0.54), feedback (ES = 0.30), pre-writing activities (ES = 0.32), and peer-assistance (ES = 0.59). As predicted, most of these approaches had smaller effects than were obtained in the less stringent methodological review by Graham et al. ([<reflink idref="bib64" id="ref140">64</reflink>]). In the prior meta-analysis, the average weighted ESs were 0.76 for strategy instruction, 0.71 for transcription instruction, 0.34 for feedback, and 0.49 for pre-writing activities. In contrast, the ESs for the process approach to writing were smaller (ES = 0.65) in Graham et al. ([<reflink idref="bib64" id="ref141">64</reflink>]) than they were in this review, as was the ES for peer-assistance (ES = 0.51). It is not clear in either case why stronger methodological studies would result in higher ESs.</p> <p>When we looked specifically at two different forms of writing strategy instruction, SRSD vs. non-SRSD (Harris & Graham, [<reflink idref="bib71" id="ref142">71</reflink>]), both of them produced statistically detectable improvements in writing when all measures were considered collectively. The ES for SRSD was 0.84, and it was 0.51 for non-SRSD. Consistent with our predictions, these ESs were smaller than the ones obtained for SRSD (0.94) and non-SRSD studies (0.66) in Graham et al. ([<reflink idref="bib64" id="ref143">64</reflink>]). The larger ESs in this and the previous review for SRSD is likely due to the fact that SRSD studies teach students procedures for regulating targeted writing strategies, the writing process, and writing behaviors as well as teaching students the knowledge needed to use the taught writing strategies and self-regulation procedures (Harris & Graham, [<reflink idref="bib71" id="ref144">71</reflink>]).</p> <p>Two especially notable outcomes in the current review were that increasing how much secondary students wrote and having them write with digital tools (word processing and word processing plus programs) did not result in a statistically detectable gain in writing performance when all writing measures were considered collectively (0.14 ES for increased writing; 0.18 ES for digital writing) or when writing quality (ES = 0.15) and ideation (ES = 0.26) were examined separately for digital writing. In the previous review by Graham et al. ([<reflink idref="bib64" id="ref145">64</reflink>]), increased writing did not enhance writing performance (ES = 0.14), but digital writing did (ES = 0.22 for word processing; ES = 0.42 for word processing plus). The findings from this and the prior meta-analysis by Graham et al. ([<reflink idref="bib64" id="ref146">64</reflink>]) suggest that writing alone has little or no effect on secondary students' writing. It is possible that writing with a computer or a computer plus program did not significantly enhance writing in the qualitatively stronger studies included in the current meta-analysis was a consequence of students not knowing how to take advantage of the affordances provided by digital writing tools. Future research needs to examine if these tools are more effective when accompanied by instruction on how to take advantage of their affordances.</p> <p>In contrast to the outcomes from Graham et al. ([<reflink idref="bib64" id="ref147">64</reflink>]), we did not find that computer-assisted writing instruction enhanced secondary students' writing performance (ES = 0.05). In the prior review, a statistically significant ES of 0.32 was obtained. We also did not obtain a statistically detectable improvement in writing when instruction in text structure was provided, even though our ES of 0.48 was larger than the statistically significant ES of 0.39 in Graham et al. ([<reflink idref="bib64" id="ref148">64</reflink>]). The most likely reason for the lack of statistical significance for text structure instruction in the present review was that the effects for individual studies were quite variable.</p> <p>The statistically significant outcomes from this review provided support for multiple aspects of the WWC model (Graham, [<reflink idref="bib57" id="ref149">57</reflink>]). We assumed that if a writing treatment was designed to improve a specific process or resource in the model (e.g., transcription skills) and there was a positive effect on writing performance, this would provide support for that aspect of the model. Consequently, the statistically significant findings for specific writing treatments provided support for the following aspects of the WWC model: production process of ideation (pre-writing activities improved writing), transcription (handwriting, spelling, and typing instruction improved writing), and conceptualization and reconceptualization (teaching writing strategies improved writing). The positive effects of writing strategy instruction also provided support for the use of schemas (or strategies) as a tool to establish executive control over some aspect of the composing process.</p> <p>We further obtained support for the hypothesized importance of social processes in the WWC model. This was supported by findings that feedback as well as having peers help each other while composing improved secondary students' writing. Because the process approach to writing draws heavily on the social aspects of composing, our finding that this writing treatment improved writing also supports the importance of social processes in writing. As additional high quality instructional research on writing emerges, evidence supporting or not supporting other aspects of the WWC and other writing models should be available.</p> <hd id="AN0184705652-36">Practical implications</hd> <p>The findings from this best evidence meta-analysis of teaching writing to secondary students yielded three important implications for classroom practices as well as policy recommendations. One, secondary students' writing can be improved by teaching writing. Unfortunately, many secondary teachers across the globe devote little time to writing instruction (Graham, [<reflink idref="bib56" id="ref150">56</reflink>]) and, as a result, students' writing is not what it should be (Graham & Rijlaarsdam, [<reflink idref="bib68" id="ref151">68</reflink>]; National Center on Educational Statistics, [<reflink idref="bib118" id="ref152">118</reflink>]). The findings from this and the prior meta-analysis by Graham et al. ([<reflink idref="bib64" id="ref153">64</reflink>]) demonstrate that this situation can be changed. It is critical, therefore, that policy makers, administrators, and teachers place greater emphasis on teaching secondary students to write.</p> <p>Two, if teachers are to improve the writing of middle and high school students, they need to do more than just have students write. In this review, increasing how much time students wrote or were engaged in using word processing or word processing plus did not enhance their writing. This does not mean that writing is unimportant. Having students write was central to most of the writing treatments we found to be effective, including the process approach to writing, writing strategy instruction, peers assisting each other when writing, and providing feedback all centered around student writing. Rather, our findings suggest that reliance on a "writing is caught" approach (Graham & Harris, [<reflink idref="bib60" id="ref154">60</reflink>]) where students just write frequently or more, with or without digital tools, is not an effective means for helping secondary students become better writers. They require instruction and guidance from educators.</p> <p>Three, the findings from this meta-analysis provided support for the contention that an effective writing program needs to embrace cognitive and social solutions for improving secondary students' writing. The positive effects of the process approach to writing, writing strategy instruction, pre-writing activities, feedback, and peers assisting each other as they write make it clear that teaching and supporting students' use of cognitive writing processes and strategies essential to composing can improve secondary students' writing. Likewise, the positive effects of transcription instruction in this review makes it clear that some secondary students still need assistance mastering basic foundational skills such as spelling. Finally, the importance of the social aspects of writing for secondary students was supported by our findings that secondary students' writing improved when they received feedback from others, worked with their peers to compose text, and were learning to write in the context of a process approach to writing (which places strong emphases on the social aspects of writing).</p> <p>It must be noted that a strong writing program cannot be based just on the writing treatments found to be effective in this best evidence meta-analysis. We provide one simple example to illustrate this. Flexible and effective sentence construction skills are absolutely central to effective writing. We can offer no advice about the teaching of such skills based on the findings from the current review because there were not enough high-quality studies assessing sentence construction writing treatments with secondary students. However, other reviews applying less stringent methodological criteria have found that teaching sentence construction has a positive effect on secondary students' writing (e.g., Graham & Perin, [<reflink idref="bib67" id="ref155">67</reflink>]; Graham et al., [<reflink idref="bib64" id="ref156">64</reflink>]). Until we have more high-quality writing intervention studies, policy makers, teacher educators, and school personnel will have to rely on both higher- and lower-quality studies to guide writing classroom practices.</p> <hd id="AN0184705652-37">Limitations and concluding comments</hd> <p>While a variety of procedures were used to locate as many relevant studies as possible, publication bias was present in this review. This should be kept in mind when interpreting our results. Because missing studies in our funnel plot were mostly from the left-side of the graph, it is possible that our overall effects for writing may be somewhat inflated.</p> <p>This review focused on just two specific types of quantitative studies: experimental and quasi-experimental investigations (with pretests). There are other types of quantitative research, such as single participant designs, that can provide useful and important information about the effectiveness of writing treatments (see Rogers & Graham, [<reflink idref="bib138" id="ref157">138</reflink>]). Investigations that involve qualitative data can also provide valuable insights into how writing is enacted in the classroom by typical and exceptional teachers. If we are to obtain a fuller picture of effective practices for teaching writing to secondary students, the findings reported in this and other meta-analyses need to be integrated with data from other reviews that examine studies applying different methodologies.</p> <p>More than one-half of the studies included in this review did not provide basic information on student characteristics such as race, gender, primary language, or socio-economic status. When this information was provided, it was very rare for researchers to provide the information needed to calculate effects for writing treatments by these student characteristics. This is a problem that has plagued the field of educational research since its inception (Graham, [<reflink idref="bib58" id="ref158">58</reflink>]). By making this issue visible, we hope that researchers will address this issue in future studies.</p> <p>Importantly, applying more stringent methodological criteria to reviewed studies as was done in the current review meant that we were able to examine just 10 writing treatments versus the 23 writing treatments examined in Graham et al. ([<reflink idref="bib64" id="ref159">64</reflink>]). Consequently, just six evidence-based writing practices were identified in this best evidence meta-analysis.</p> <p>Finally, we limited our inclusion of studies to those included in the meta-analysis conducted by Graham et al. ([<reflink idref="bib64" id="ref160">64</reflink>]). The search for this previous meta-analysis ended September 30, 2021. This means that studies conducted since that date were not included in this best evidence meta-analysis.</p> <p>In conclusion, this best evidence meta-analysis demonstrated that secondary students' writing can be improved when they are taught how to write. Just as importantly, we identified multiple writing treatments that can improve the writing of these adolescents. Consequently, the findings from this review and Graham et al. ([<reflink idref="bib64" id="ref161">64</reflink>]) provide fertile information that policy makers, administrators, teacher educators, and teachers can use to improve writing instruction for middle and high school students.</p> <hd id="AN0184705652-38">Acknowledgements</hd> <p>The research reported here was supported by the Institute of Education Sciences, U.S. Department of Education, through Grant R305C190007 to the University of California – Irvine for the WRITE Center. The opinions expressed are those of the authors and do not represent views of the Institute or the U.S. Department of Education.</p> <hd id="AN0184705652-39">Declarations</hd> <p></p> <hd id="AN0184705652-40">Conflict of interest</hd> <p>None of the authors have a conflict of interest.</p> <hd id="AN0184705652-41">Publisher's Note</hd> <p>Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.</p> <ref id="AN0184705652-42"> <title> References </title> <blist> <bibl id="bib1" idref="ref62" type="bt">1</bibl> <bibtext> *Adams, V, (1971). A study of the effects of two methods of teaching composition to twelfth Graders [Unpublished doctoral dissertation]. University of Illinois at Urbana-Champaign.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref37" type="bt">2</bibl> <bibtext> Aiken L, West S, Schwalm D, Carroll J, Hsiung S. Comparison of a randomized and two quasi-experimental designs in a single outcome evaluation. Evaluation Review. 1998; 22: 207-244. 10.1177/0193841X9802200203</bibtext> </blist> <blist> <bibl id="bib3" idref="ref64" type="bt">3</bibl> <bibtext> *Al Shaheb, M. N. A. (n.d.). The effect of self-regulated strategy development on persuasive writing, self-efficacy, and attitude: A mixed-methods, quasi-experimental study in Grade 6 in Lebanon [Unpublished doctoral dissertation]. Université Saint-Joseph, Beirut, Lebanon.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref11" type="bt">4</bibl> <bibtext> Applebee A, Langer J. A snapshot of writing instruction in middle schools and high schools. English Journal. 2011; 100: 14-27. 10.58680/ej201116413</bibtext> </blist> <blist> <bibl id="bib5" idref="ref17" type="bt">5</bibl> <bibtext> Bangert-Drowns R. The word processor as an instructional tool: A meta-analysis of word processing in writing instruction. Review of Educational Research. 1993; 63: 69-93. 10.3102/00346543063001069</bibtext> </blist> <blist> <bibl id="bib6" idref="ref1" type="bt">6</bibl> <bibtext> Bangert-Drowns R, Hurley M, Wilkinson B. The effects of school-based writing to-learn interventions on academic achievement: A meta-analysis. Review of Educational Research. 2004; 74; 1: 29-58. 10.3102/00346543074001029</bibtext> </blist> <blist> <bibl id="bib7" idref="ref68" type="bt">7</bibl> <bibtext> Barrot JS. Using the sociocognitive-transformative approach in writing classrooms: Effects on L2 learners' writing performance. Reading & Writing Quarterly. 2018; 34; 2: 187-201. 10.1080/10573569.2017.1387631</bibtext> </blist> <blist> <bibl id="bib8" type="bt">8</bibl> <bibtext> *Barton, H. (2018). Writing, collaborating, and cultivating: Building writing self-efficacy and skills through a student-centric, student-led writing center. Doctor of Education in Secondary Education Dissertations. 13. https://digitalcommons.kennesaw.edu/seceddoc_etd/13</bibtext> </blist> <blist> <bibl id="bib9" type="bt">9</bibl> <bibtext> *Benson, N. L. (1979). The effects of peer feedback during the writing process on writing performance, revision behavior, and attitude toward writing [Unpublished doctoral dissertation] University of Colorado.</bibtext> </blist> <blist> <bibtext> Berman R. Learners' transfer of writing skills between languages. TESL Canada Journal. 1994; 12; 1: 29-46. 10.18806/tesl.v12i1.642</bibtext> </blist> <blist> <bibtext> *Black, J. G. (1995). Teaching elements of written composition through use of classical music and art: the effects on high school students' writing [Unpublished doctoral dissertation] University of California, Riverside.</bibtext> </blist> <blist> <bibtext> Bloom, B, Engelhart, M, Furst, E, Hill, W, & Krathwohl, D. (1956). Taxonomy of educational objectives: The classification of educational goals. Handbook I: Cognitive domain. David McKay Company.</bibtext> </blist> <blist> <bibtext> *Braaksma, M. (2002). Observational learning in argumentative writing [Unpublished doctoral dissertation]. Amsterdam: University of Amsterdam.</bibtext> </blist> <blist> <bibtext> *Braaksma, M, Rijlaarsdam, G. C. W, & van den Bergh, H. H. (2018). Effects of hypertext writing and observational learning on content knowledge acquisition, self-efficacy, and text quality: Two experimental studies exploring aptitude treatment interactions. Journal of Writing Research, 9(3), 259–300. https://doi.org/10.17239/jowr-2018.09.03.02</bibtext> </blist> <blist> <bibtext> Brantley H, Small D. Effects of self evaluation on written composition skill in learning disabled children. 1991; U.S. Department of Education</bibtext> </blist> <blist> <bibtext> *Brewer, D. (2002). Teaching writing in science through the use of a writing rubric [Unpublished doctoral dissertation]. University of Michigan-Flint.</bibtext> </blist> <blist> <bibtext> Cheung A, Slavin R. How methodological features affect effect sizes in education. Educational Researcher. 2016; 45: 283-292. 10.3102/0013189X16656615</bibtext> </blist> <blist> <bibtext> Christensen CA. Relationship between orthographic-motor integration and computer use for the production of creative and well-structured written text. British Journal of Educational Psychology. 2004; 74; 4: 551-564. 10.1348/0007099042376373</bibtext> </blist> <blist> <bibtext> Chung HQ, Chen V, Olson CB. The impact of self-assessment, planning and goal setting, and reflection before and after revision on student self-efficacy and writing performance. Reading and Writing. 2021; 34: 1885-1913. 10.1007/s11145-021-10186-x</bibtext> </blist> <blist> <bibtext> Combs WE. Further effects of sentence-combining practice on writing ability. Research in the Teaching of English. 1976; 10; 2: 137-149. 10.58680/rte197620035</bibtext> </blist> <blist> <bibtext> Combs WE. Sentence-combining practice: Do gains in judgments of writing "quality" persist?. The Journal of Educational Research. 1977; 70; 6: 318-321. 10.1080/00220671.1977.10885014</bibtext> </blist> <blist> <bibtext> *Conklin, E. (2007). Concept mapping: Impact on content and organization of technical writing in science [Unpublished doctoral dissertation]. Walden University.</bibtext> </blist> <blist> <bibtext> *Corey, D. R. (1990). The effects of concurrent instruction in composition and speech upon the composition writing holistic scores and sense of audience of a ninth-grade student population [Unpublished doctoral dissertation]. The Union for Experimenting Colleges and Universities.</bibtext> </blist> <blist> <bibtext> Cortina J, Nouri H. Effect size for ANOVA design. 2000; Sage. 10.4135/9781412984010</bibtext> </blist> <blist> <bibtext> *Couzijn, M, & Rijlaarsdam, G. (2004). Learning to read and write argumentative text by observation of peer learners. In Rijlaarsdam, G. (Series Ed.) & Rijlaarsdam, G, van den Bergh, H, & Couzijn, M. (Vol. 14 Eds.), Effective learning and teaching of writing, (pp. 241–258). Springer.</bibtext> </blist> <blist> <bibtext> Couzijn M, Rijlaarsdam GRijlaarsdam G, van den Bergh H, Couzijn M. Learning to write instructive texts by reader observation and written feedback. Effective learning and teaching of writing. 2005; Springer: 209-240. 10.1007/978-1-4020-2739-0_16</bibtext> </blist> <blist> <bibtext> *Covill, A. E. (1996). Students' revision practices and attitudes in response to surface-related feedback as compared to content-related feedback on their writing [Unpublished doctoral dissertation] University of Washington.</bibtext> </blist> <blist> <bibtext> Cremin T, Myhill D, Eyres I, Nash T, Wilson A, Oliver L. Teachers as writers: Learning together with others. Literacy. 2020; 54; 2: 49-59. 10.1111/lit.12201</bibtext> </blist> <blist> <bibtext> *Crook, J. D. (1985). Effects of computer-assisted instruction upon seventh-grade students' growth in writing performance [Unpublished doctoral dissertation]. University of Nebraska.</bibtext> </blist> <blist> <bibtext> *Crossley, S. A, Roscoe, R, & McNamara, D. S. (2013). Using automated scoring models to detect changes in student writing in an intelligent tutoring system. Paper presented at Proceedings of the Twenty-sixth International Florida Artificial Intelligence Research Society Conference, Florida.</bibtext> </blist> <blist> <bibtext> *Crossley, S. A, Varner, L. K, Roscoe, R. D, & McNamara, D. S. (2013). Using automated indices of cohesion to evaluate an intelligent tutoring system and an automated writing evaluation system. In Artificial intelligence in education: 16th international conference, AIED 2013, Memphis, TN, USA, July 9–13, 2013. Proceedings 16 (pp. 269-278). Springer, Berlin, Heidelberg.</bibtext> </blist> <blist> <bibtext> *Dailey, E. M. (1992). The relative efficacy of cooperative learning versus individualized learning on the written performance of adolescent students with writing problems [Unpublished doctoral dissertation]. The Johns Hopkins University.</bibtext> </blist> <blist> <bibtext> Daiute C, Kruidenier J. A self-questioning strategy to increase young writers' revising processes. Applied Psycholinguistics. 1985; 6: 307-318. 10.1017/S0142716400006226</bibtext> </blist> <blist> <bibtext> de la Paz S, Graham S. Explicitly teaching strategies, skills, and knowledge: Writing instruction in middle school classrooms. Journal of Educational Psychology. 2002; 94; 4: 687. 10.1037/00220663944687</bibtext> </blist> <blist> <bibtext> *de la Paz, S, & Wissinger, D. R. (2017). Improving the historical knowledge and writing of students with or at risk for LD. Journal of Learning Disabilities, 50(6), 658–671. https://doi.org/10.1177/0022219416659444</bibtext> </blist> <blist> <bibtext> *de la Paz, S, Wissinger, D. R, Gross, M, & Butler, C. (in press). Strategies that promote historical reasoning and contextualization: A pilot intervention with urban high school students. Reading and Writing.</bibtext> </blist> <blist> <bibtext> *de Ment, L. (2008). The Relationship of self-evaluation, writing ability, and attitudes toward writing among gifted grade 7 language arts students [Unpublished master's thesis] Walden University.</bibtext> </blist> <blist> <bibtext> de Smedt F, van Keer H. Fostering writing in upper primary grades: A study into the distinct and combined impact of explicit instruction and peer assistance. Reading and Writing. 2018; 31: 325-354. 10.1007/s11145-017-9787-4</bibtext> </blist> <blist> <bibtext> de Smedt F, Graham S, van Keer H. The bright and dark side of writing motivation: Effects of explicit instruction and peer assistance. The Journal of Educational Research. 2019; 112; 2: 152-167. 10.1080/00220671.2018.1461598</bibtext> </blist> <blist> <bibtext> de Smedt F, Graham S, Van Keer H. "It takes two": The added value of structured peer-assisted writing in explicit writing instruction. Contemporary Educational Psychology. 2020; 60: 101835. 10.1016/j.cedpsych.2019.101835</bibtext> </blist> <blist> <bibtext> Drew S, Olinghouse N, Luby-Faggella M, Welsh M. Framework for disciplinary writing in science Grades 6–12: A national survey. Journal of Educational Psychology. 2017; 109: 935-955. 10.1037/edu0000186</bibtext> </blist> <blist> <bibtext> Dunnagan, K. L. (1990). Seventh grade students' audience awareness in writing produced within and without the dramatic mode [Unpublished doctoral dissertation]. The Ohio State University.</bibtext> </blist> <blist> <bibtext> *Eliason, R. G. (1994). The effect of selected word processing adjunct programs on the writing of high school students (Publication No. 0426055) [Doctoral dissertation]. University of South Florida.</bibtext> </blist> <blist> <bibtext> *Erickson, D. K. (2009). The effects of blogs versus dialogue journals on open-response writing scores and attitudes of grade eight science students (Publication No. 3393920) [Doctoral dissertation]. University of Massachusetts, Lowell. ProQuest LLC.</bibtext> </blist> <blist> <bibtext> *Espinoza, S. F. (1992). The effects of using a word processor containing grammar and spell checkers on the composition writing of sixth graders [Unpublished doctoral dissertation]. Texas Tech University.</bibtext> </blist> <blist> <bibtext> Festas I, Oliveira AL, Rebelo JA, Damião MH, Harris K, Graham S. Professional development in self-regulated strategy development: Effects on the writing performance of eighth grade Portuguese students. Contemporary Educational Psychology. 2015; 40: 17-27. 10.1016/j.cedpsych.2014.05.004</bibtext> </blist> <blist> <bibtext> Fisher, Z, Tipton, E, & Zhipeng, H. (2017). Package ' robumeta'. Retrieved from <ulink href="http://cran.uni-muenster.de/web/packages/robumeta/robumeta.pdf">http://cran.uni-muenster.de/web/packages/robumeta/robumeta.pdf</ulink></bibtext> </blist> <blist> <bibtext> *Frank, A. R. (2008). The effect of instruction in orthographic conventions and morphological features on the reading fluency and comprehension skills of high-school freshmen [Unpublished doctoral dissertation]. The University of San Francisco.</bibtext> </blist> <blist> <bibtext> *Franzke, M, Kintsch, E, Caccamise, D, Johnson, N, & Dooley, S. (2005). Summary Street®: Computer support for comprehension and writing. Journal of Educational Computing Research, 33(1), 53–80. https://doi.org/10.2190/DH8F-QJWM-J457-FQVB</bibtext> </blist> <blist> <bibtext> *Frost, K. L. (2008). The effects of automated essay scoring as a high school classroom intervention (Publication No. 3352171) [Doctoral dissertation], University of Nevada, Las Vegas. ProQuest LLC.</bibtext> </blist> <blist> <bibtext> *Galbraith, J. (2014). The effect of self-regulation writing strategies and gender on writing self-efficacy and persuasive writing achievement for secondary students [Unpublished doctoral dissertation]. Western Connecticut State University.</bibtext> </blist> <blist> <bibtext> *Ganong, F. L. (1974). Teaching writing through the use of a program based on the work of Donald M. Murray [Unpublished doctoral dissertation]. Boston University.</bibtext> </blist> <blist> <bibtext> Goldberg, A, Russell, M, & Cook, A. (2003). The effect of computers on student writing: A meta-analysis of studies from 1992 to 2002. The Journal of Technology, Learning, and Assessment2(1). Retrived from https://ejournals.bc.edu/index.php/jtla/article/view/1661</bibtext> </blist> <blist> <bibtext> González-Lamas J, Cuevas I, Mateos M. Arguing from sources: Design and evaluation of a programme to improve written argumentation and its impact according to students' writing beliefs. Journal for the Study of Education and Development. 2016; 39; 1: 49-83. 10.1080/02103702.2015.111160</bibtext> </blist> <blist> <bibtext> *Grejda, G. F. (1988). The effects of word processing and revision patterns on the writing quality of sixth-grade students (Publication No. 8909998) [Doctoral dissertation], Pennsylvania State University.</bibtext> </blist> <blist> <bibtext> Graham S. Changing how writing is taught. Review of Research in Education. 2019; 43: 277-303. 10.3102/0091732X18821125</bibtext> </blist> <blist> <bibtext> Graham S. A revised writer(s)-within-community model of writing. Educational Psychologist. 2018; 53: 258-279. 10.1080/00461520.2018.1481406</bibtext> </blist> <blist> <bibtext> Graham S. Inaugural editorial for the journal of educational psychology. Journal of Educational Psychology. 2015; 107: 1-2. 10.1037/edu0000007</bibtext> </blist> <blist> <bibtext> Graham, S, & Harris, K. R. (2018). Evidence-based writing practices: A meta-analysis of existing meta-analyses. In R. Fidalgo, K. R. Harris, & M. Braaksma (Eds.). Design principles for teaching effective writing: Theoretical and empirical grounded principles (pp. 13–37). Hershey, PA: Brill Editions.</bibtext> </blist> <blist> <bibtext> Graham S, Harris KR. It can be taught, but it does not develop naturally: Myths and realities in writing instruction. School Psychology Review. 1997; 26: 414-424. 10.1080/02796015.1997.12085875</bibtext> </blist> <blist> <bibtext> Graham, S, & Harris, K. R. (2014). Conducting high quality writing intervention research: Twelve recommendations. Journal of Writing Research 6, 89–123. https://doi.org/10.17239/jowr-2014.06.02.1</bibtext> </blist> <blist> <bibtext> Graham, S, & Hebert, M. (2011). Writing-to-read: A meta-analysis of the impact of writing and writing instruction on reading. Harvard Educational Review, 81, 710–744. https://doi.org/10.17763/haer.81.4.t2k0m13756113566</bibtext> </blist> <blist> <bibtext> Graham S, Hebert M, Harris KR. Formative assessment and writing: A meta-analysis. Elementary School Journal. 2015; 115: 524-547. 10.1086/681947</bibtext> </blist> <blist> <bibtext> Graham S, Kim Y, Cao Y, Lee W, Tate T, Collins T, Cho M, Moon Y, Chung H, Olson C. A meta-analysis of writing treatments for students in Grades 6 to 12. Journal of Educational Psychology. 2023; 115: 1004-1027. 10.1037/edu0000819</bibtext> </blist> <blist> <bibtext> Graham S, Kiuhara S, MacKay M. The effects of writing on learning in science, social studies, and mathematics: A meta-analysis. Review of Educational Research. 2020; 90: 179-226. 10.3102/0034654320914744</bibtext> </blist> <blist> <bibtext> Graham S, Kiuhara S, McKeown D, Harris KR. A meta-analysis of writing instruction for students in the elementary grades. Journal of Educational Psychology. 2012; 104: 879-896. 10.1037/a0029185</bibtext> </blist> <blist> <bibtext> Graham S, Perin D. A meta-analysis of writing instruction for adolescent students. Journal of Educational Psychology. 2007; 99: 445-476. 10.1037/0022-0663.99.3.445</bibtext> </blist> <blist> <bibtext> Graham S, Rijlaarsdam G. Writing education around the globe: Introduction and call for a new global analysis. Reading & Writing: An Interdisciplinary Journal. 2016; 29: 781-792. 10.1007/s11145-016-9640-1</bibtext> </blist> <blist> <bibtext> *Hamilton, H. (1960). A combined auditory-visual syllabification approach to the study of spelling [Unpublished master's thesis]. Texas Technological College.</bibtext> </blist> <blist> <bibtext> *Hammar, D. D. (1986). The effectiveness of computer-assisted writing instruction for juniors who have failed the Regents competency test in writing (Publication No. 8704355) [Doctoral dissertation]. University of Rochester.</bibtext> </blist> <blist> <bibtext> Harris KR, Graham SPressley M, Harris KR, Guthrie J. Self-regulated strategy development: A part of the writing process. Promoting academic competence and literacy in school. 1992; Academic Press: 277-309</bibtext> </blist> <blist> <bibtext> *Harville, M. L. (2001). A study of computer-assisted expository writing of middle school students with special learning needs [Unpublished doctoral dissertation]. Columbia University</bibtext> </blist> <blist> <bibtext> Hedges LV, Tipton E, Johnson MC. Robust variance estimation in meta-regression with dependent effect size estimates. Research Synthesis Methods. 2010; 1; 1: 39-65. 10.1002/jrsm.5</bibtext> </blist> <blist> <bibtext> *Hickerson, B. L. (1987). Critical thinking, reading, and writing: developing a schema for expository text through direct instruction in analysis of text structure (metacognition, webbing, mapping) [Unpublished doctoral dissertation]. North Texas State University.</bibtext> </blist> <blist> <bibtext> *Higgins, P. D. (2013). The effects of using a critical thinking graphic organizer to improve Connecticut academic performance test interdisciplinary writing assessment scores [Unpublished doctoral dissertation]. Western Connecticut State University.</bibtext> </blist> <blist> <bibtext> *Hill, B. G. (1990). A comparison of the writing quality of paired and unpaired students composing at the computer [Unpublished doctoral dissertation]. University of Texas at Austin.</bibtext> </blist> <blist> <bibtext> Hillocks GBazerman C. Writing in secondary schools. Handbook of research on writing: History, society, school, individual, text. 2008; Routledge: 311-329</bibtext> </blist> <blist> <bibtext> Hillocks, G. J. (1986). Research on written composition: New directions for teaching. National Council of Teachers of English.</bibtext> </blist> <blist> <bibtext> Hillocks G. The interaction of instruction, teacher comment, and revision in teaching the composing process. Research in the Teaching of English. 1982; 16; 3: 261-278. 10.58680/rte198215736</bibtext> </blist> <blist> <bibtext> Hisgen S, Barwasser A, Wellmann T, Grünke M. The effects of a multicomponent strategy instruction on the argumentative writing performance of low-achieving secondary students. Learning Disabilities: A Contemporary Journal. 2020; 18; 1: 93-110</bibtext> </blist> <blist> <bibtext> Higgins, J. P. T, & Green, S. (Eds.). (2011). Cochrane handbook for systematic reviews of interventions (Version 5.1.0). Retrieved from <ulink href="http://handbook.cochrane.org">http://handbook.cochrane.org</ulink></bibtext> </blist> <blist> <bibtext> *Holley, C. A. B. (1990). The effects of peer editing as an instructional method on the writing proficiency of selected high school students in Alabama [Unpublished doctoral dissertation]. University of Alabama.</bibtext> </blist> <blist> <bibtext> Holliway D, McCutchen DAllal L, Chanquoy L, Largy P. Audience perspective in young writers' composing and revising. Revision: Cognitive and instructional processes. 2004; Kluwer: 87-101. 10.1007/978-94-007-1048-1_6</bibtext> </blist> <blist> <bibtext> *Hoogeveen, M. C. E. J. (2013). Writing with peer response using genre knowledge: A classroom intervention study [Unpublished doctoral dissertation]. University of Twente.</bibtext> </blist> <blist> <bibtext> Hoogeveen M, van Gelderen A. Effects of peer response using genre knowledge on writing quality: A randomized control trial. The Elementary School Journal. 2015; 116; 2: 265-290. 10.1086/684129</bibtext> </blist> <blist> <bibtext> Hoogeveen M, van Gelderen A. Writing with peer response using different types of genre knowledge: Effects on linguistic features and revisions of sixth-grade writers. The Journal of Educational Research. 2018; 111; 1: 66-80. 10.1080/00220671.2016.1190913</bibtext> </blist> <blist> <bibtext> Hopewell S, Loudon K, Clarke MJ, Oxman AD, Dickersin K. Publication bias in clinical trials due to statistical significance or direction of trial results. Cochrane Database of Systematic Reviews. 2009; 1: 1-25. 10.1002/14651858.MR000006.pub3</bibtext> </blist> <blist> <bibtext> Iordanou K, Constantinou CP. Supporting use of evidence in argumentation through practice in argumentation and reflection in the context of SOCRATES learning environment. Science Education. 2015; 99; 2: 282-311. 10.1002/sce.21152</bibtext> </blist> <blist> <bibtext> *Jacoby, K. E. (1990). Remove the dust covers and let the children play: An investigation into the effectiveness of computers in spelling drill and practice in the classroom [Unpublished master's thesis]. University of New England.</bibtext> </blist> <blist> <bibtext> *Jeroski, S. (1982). Competence in written expression: Interactions between instruction and individual differences among junior high school students [Unpublished doctoral dissertation]. University of British Columbia.</bibtext> </blist> <blist> <bibtext> *Jones, J. L. (1966). Effects of spelling instruction in eighth grade biological science upon scientific spelling, vocabulary, and reading comprehension; general spelling, vocabulary and reading comprehension; science progress; and science achievement [Unpublished doctoral dissertation]. University of Maryland.</bibtext> </blist> <blist> <bibtext> Jones S, Myhill D, Bailey T. Grammar for writing? An investigation of the effects of contextualised grammar teaching on students' writing. Reading and Writing. 2013; 26; 8: 1241-1263. 10.1007/s11145-012-9416-1</bibtext> </blist> <blist> <bibtext> *Kaffar, B. (1993). Exploring the effects of online instructional models on the writing achievement of high school students with and without disabilities [Unpublished doctoral dissertation]. University of Nevada.</bibtext> </blist> <blist> <bibtext> *Kasparek, R. F. (1994). Effects of integrated writing on attitude and algebra performance of high school students [Unpublished doctoral dissertation]. The University of North Carolina at Greensboro.</bibtext> </blist> <blist> <bibtext> *Kelley, K. R. (1984). The effect of writing instruction on reading comprehension and story writing ability [Unpublished doctoral dissertation]. University of Pittsburgh.</bibtext> </blist> <blist> <bibtext> *Kennedy, K. A. (2008). Validating FOLA: A randomized writing experiment [Unpublished doctoral dissertation]. University of Southern California.</bibtext> </blist> <blist> <bibtext> Kim JS, Olson CB, Scarcella R, Kramer J, Pearson M, van Dyk D, Collins P, Land RE. A randomized experiment of a cognitive strategies approach to text-based analytical writing for mainstreamed Latino English language learners in grades 6 to 12. Journal of Research on Educational Effectiveness. 2011; 4; 3: 231-263. 10.1080/19345747.2010.523513</bibtext> </blist> <blist> <bibtext> Koster M, Tribushinina E, de Jong PF, van den Bergh H. Teaching children to write: A meta-analysis of writing intervention research. Journal of Writing Research. 2015; 7; 2: 249-274. 10.17239/jowr-2015.07.02.2</bibtext> </blist> <blist> <bibtext> *Kuscenko, D. (2018). Supporting collaborative writing in secondary Language Arts: A revision decision method intervention [Unpublished doctoral dissertation]. Lehigh University.</bibtext> </blist> <blist> <bibtext> *Lane, R. A. J. (2003). "Keep cool and DECIDE": Using discussion and writing instruction to improve the problem-solving skills of adolescent, urban students, labeled learning disabled [Unpublished doctoral dissertation]. Columbia University.</bibtext> </blist> <blist> <bibtext> Lange AA, Mulhern G, Wylie J. Proofreading using an assistive software homophone tool. Journal of Learning Disabilities. 2009; 42; 4: 322-335. 10.1177/0022219408331035</bibtext> </blist> <blist> <bibtext> *Laysears-Smith, R. R. (2005). Students' attitudes about structured journal writing and their perceptions of their self-esteem in an urban career and technical classroom [Unpublished doctoral dissertation]. Temple University.</bibtext> </blist> <blist> <bibtext> Lee J, Schallert DL. Exploring the reading-writing connection: A yearlong classroom-based experimental study of middle school students developing literacy in a new language. Reading Research Quarterly. 2016; 51; 2: 143-164. 10.1002/rrq.132</bibtext> </blist> <blist> <bibtext> Limpo T, Alves RA. Implicit theories of writing and their impact on students' response to a SRSD intervention. British Journal of Educational Psychology. 2014; 84; 4: 571-590. 10.1111/bjep.12042</bibtext> </blist> <blist> <bibtext> Lipskey M, Wilson D. Practical meta-analysis. 2001; Sage</bibtext> </blist> <blist> <bibtext> López P, Torrance M, Rijlaarsdam G, Fidalgo R. Evaluating effects of different forms of revision instruction in upper-primary students. Reading and Writing. 2021; 34: 1741-1767. 10.1007/s11145-021-10156-3</bibtext> </blist> <blist> <bibtext> *Lott, C. J. (1986). The effects of the microcomputer word processor on the composition skills of seventh-grade students [Unpublished doctoral dissertation]. University of Montana.</bibtext> </blist> <blist> <bibtext> *Lyons, H. L. K. (2002). The effects of technology use on student writing proficiency and student attitudes toward written assignments in a ninth-grade language arts classroom [Unpublished doctoral dissertation]. Idaho State University.</bibtext> </blist> <blist> <bibtext> *Lytle, M. J. (1987). Word processors and writing: The relation of seventh grade students' learner characteristics and revision behaviors (Publication No. 8800537) [Doctoral dissertation], University of Oregon.</bibtext> </blist> <blist> <bibtext> Matuchniak T, Olson CB, Scarcella R. Examining the text-based, on-demand, analytical writing of mainstreamed Latino English learners in a randomized field trial of the Pathway Project intervention. Reading and Writing. 2014; 27: 973-994. 10.1007/s11145-013-9490-z</bibtext> </blist> <blist> <bibtext> *Mayo, N. B. (1976). The effects of discussion and assignment questions on the quality of descriptive writing of tenth grade students [Unpublished doctoral dissertation]. Memphis State University.</bibtext> </blist> <blist> <bibtext> *McCarty, R. P. (2016). Leveraging historical thinking heuristics as warrants in historical argumentative writing [Unpublished doctoral dissertation]. University of Illinois at Chicago.</bibtext> </blist> <blist> <bibtext> *McCreight, C. K. (1995). Computer-assisted process writing: A cooperative-pairs approach [Unpublished doctoral dissertation]. Baylor University.</bibtext> </blist> <blist> <bibtext> *McDermott, M. A. (2009). The impact of embedding multiple modes of representation on student construction of chemistry knowledge [Unpublished doctoral dissertation]. The University of Iowa.</bibtext> </blist> <blist> <bibtext> McNeill KL, Lizotte DJ, Krajcik J, Marx RW. Supporting students' construction of scientific explanations by fading scaffolds in instructional materials. The Journal of the Learning Sciences. 2006; 15; 2: 153-191. 10.1207/s15327809jls1502_1</bibtext> </blist> <blist> <bibtext> Morphy P, Graham S. Word processing programs and weaker writers/readers: A meta-analysis of research findings. Reading and Writing: An Interdisciplinary Journal. 2012; 25: 641-678. 10.1007/s11145-010-9292-5</bibtext> </blist> <blist> <bibtext> *Moseley, D. S. (2003). Vocabulary instruction and its effects on writing quality [Unpublished doctoral dissertation]. Louisiana Tech University.</bibtext> </blist> <blist> <bibtext> National Center for Educational Statistics (2012). The nation's report card: Writing 2011 (NCES 2012-470). Washington, DC: U.S. Department of Education, Institute of Educational Sciences.</bibtext> </blist> <blist> <bibtext> Niemi D, Wang J, Steinberg DH, Baker EL, Wang H. Instructional sensitivity of a complex language arts performance assessment. Educational Assessment. 2007; 12; 3–4: 215-237. 10.1080/10627190701578271</bibtext> </blist> <blist> <bibtext> Nunnally J. Psychometric theory. 2017; McGraw-Hill</bibtext> </blist> <blist> <bibtext> Olina Z, Sullivan HJ. Student self-evaluation, teacher evaluation, and learner performance. Educational Technology Research and Development. 2004; 52; 3: 5-22. 10.1007/BF02504672</bibtext> </blist> <blist> <bibtext> Olson CB, Kim JS, Scarcella R, Kramer J, Pearson M, van Dyk DA, Collins P, Land RE. Enhancing the interpretive reading and analytical writing of mainstreamed English learners in secondary school: Results from a randomized field trial using a cognitive strategies approach. American Educational Research Journal. 2012; 49; 2: 323-355. 10.3102/0002831212439434</bibtext> </blist> <blist> <bibtext> Olson CB, Matuchniak T, Chung HQ, Stumpf R, Farkas G. Reducing achievement gaps in academic writing for Latinos and English learners in Grades 7–12. Journal of Educational Psychology. 2017; 109; 1: 1-21. 10.1037/edu0000095</bibtext> </blist> <blist> <bibtext> Page-Voth V, Graham S. Effects of goal setting and strategy use on the writing performance and self-efficacy of students with writing and learning problems. Journal of Educational Psychology. 1999; 91; 2: 230-240. 10.1037/0022-0663.91.2.230</bibtext> </blist> <blist> <bibtext> Palumbo DB, Prater DL. A comparison of computer-based prewriting strategies for basic ninth-grade writers. Computers in Human Behavior. 1992; 8; 1: 63-70. 10.1016/0747-5632(92)90019-B</bibtext> </blist> <blist> <bibtext> *Pedersen, E. L. (1977). Improving syntactic and semantic fluency in writing of language arts students through extended practice in sentence combining [Unpublished doctoral dissertation]. The University of Minnesota.</bibtext> </blist> <blist> <bibtext> *Pittman, R. T. (2007). Improving spelling ability among speakers of African American vernacular English: An intervention based on phonological, morphological, and orthographic principle [Unpublished doctoral dissertation]. Texas A&M University.</bibtext> </blist> <blist> <bibtext> *Pivarnik, B. A. (1985). The effect of training in word processing on the writing of eleventh grade students [Unpublished doctoral dissertation]. University of Connecticut.</bibtext> </blist> <blist> <bibtext> Prata MJ, de Sousa B, Festas I, Oliveira AL. Cooperative methods and self-regulated strategies development for argumentative writing. The Journal of Educational Research. 2019; 112; 1: 12-27. 10.1080/00220671.2018.1427037</bibtext> </blist> <blist> <bibtext> Pressley M, Graham S, Harris KR. The state of educational intervention research. British Journal of Educational Psychology. 2006; 76: 1-19. 10.1348/000709905x66035</bibtext> </blist> <blist> <bibtext> *Rapanta, C. (2021). Can teachers implement a student-centered dialogical argumentation method across the curriculum? Manuscript submitted for publication.</bibtext> </blist> <blist> <bibtext> R Core Team (2018). R: A language and environment for statistical computing. Retrieved from https://<ulink href="http://www.r-project.org/">www.r-project.org/</ulink></bibtext> </blist> <blist> <bibtext> Reynolds CJ, Hill DS, Swassing RH, Ward ME. The effects of revision strategy instruction on the writing performance of students with learning disabilities. Journal of Learning Disabilities. 1988; 21; 9: 540-545. 10.1177/002221948802100904</bibtext> </blist> <blist> <bibtext> Reynolds GA, Perin D. A comparison of text structure and self-regulated writing strategies for composing from sources by middle school students. Reading Psychology. 2009; 30; 3: 265-300. 10.1080/02702710802411547</bibtext> </blist> <blist> <bibtext> *Rice, D. P. (1968). A study of a linguistically-based spelling program in grade six [Unpublished doctoral dissertation]. Temple University.</bibtext> </blist> <blist> <bibtext> *Rijlaarsdam, G, & Schoonen, R. (1988). Effects of a teaching program based on peer evaluation on written composition and some variables related to writing apprehension. SCO Cahier Nr. 47. Stichting Centrum voor Onderwijsonderzoek (SCO), Grote Bickersstraat 72, 1013 KS Amsterdam, The Netherlands.</bibtext> </blist> <blist> <bibtext> Rijlaarsdam G, Couzijn M, Janssen T, Braaksma M, Kieft M. Writing experiment manuals in science education: The impact of writing, genre, and audience. International Journal of Science Education. 2006; 28; 2–3: 203-233. 10.1080/09500690500336932</bibtext> </blist> <blist> <bibtext> Rogers L, Graham S. A meta-analysis of single subject design writing intervention research. Journal of Educational Psychology. 2008; 100: 879-906. 10.1037/0022-0663.100.4.879</bibtext> </blist> <blist> <bibtext> *Rolfe, A. B. (1991). The effects of an identify-generate-test sequence on the spelling performance of learning disabled and normally-achieving students (Publication No. 0121204) [Doctoral dissertation]. Columbia University.</bibtext> </blist> <blist> <bibtext> Roscoe RD, Allen LK, McNamara DS. Contrasting writing practice formats in a writing strategy tutoring system. Journal of Educational Computing Research. 2019; 57; 3: 723-754. 10.1177/0735633118763429</bibtext> </blist> <blist> <bibtext> Roscoe RD, Brandon R, Snow E, McNamara DSPytash K, Ferdig R. Game-based writing strategy practice with the Writing Pal. Exploring technology for writing and writing instruction. 2013; IGI Global: 1-20</bibtext> </blist> <blist> <bibtext> Rosenbluth GS, Reed WM. The effects of writing-process-based instruction and word processing on remedial and accelerated 11th graders. Computers in Human Behavior. 1992; 8: 71-95. 10.1016/0747-5632(92)90020-F</bibtext> </blist> <blist> <bibtext> RStudio Team (2016). RStudio: Integrated Development for R. Retrieved from <ulink href="http://www.rstudio.com/">http://www.rstudio.com/</ulink>.</bibtext> </blist> <blist> <bibtext> Sandmel K, Graham S. The process writing approach: A meta-analysis. Journal of Educational Research. 2011; 104: 396-407. 10.1080/00220671.2010.488703</bibtext> </blist> <blist> <bibtext> Schulz K, Grimes D. Sample size slippages in randomized trials: Exclusions and the lost and wayward. Lancet. 2002; 359: 781-785. 10.1016/s0140-6736(02)07882-0</bibtext> </blist> <blist> <bibtext> Segers E, Verhoeven L. Learning in a sheltered Internet environment: The use of WebQuests. Learning and Instruction. 2009; 19: 423-432. 10.1016/j.learninstruc.2009.02.017</bibtext> </blist> <blist> <bibtext> Slavin RE, Madden NA. Measures inherent to treatments in program effectiveness reviews. Journal of Research on Educational Effectiveness. 2011; 4: 370-380. 10.1080/19345747.2011.558986</bibtext> </blist> <blist> <bibtext> *Sloan, C. C. (2017). Types of feedback in peer review and the effect on student motivation and writing quality (Publication No. 10281143) [Doctoral dissertation]. Michigan State University. ProQuest LLC.</bibtext> </blist> <blist> <bibtext> *Spilton, R. (1986). The effects of individualized language arts, sentence-combining, and traditional grammar on the syntactic maturity and quality of writing of a select group of eighth graders (Publication No. 8703964) [Doctoral dissertation]. Georgia State University.</bibtext> </blist> <blist> <bibtext> Swanson, L, Harris, K. R, & Graham, S. (2013). Handbook of learning disabilities (Second Edition). Guilford.</bibtext> </blist> <blist> <bibtext> Tanner-Smith EE, Tipton E, Polanin JR. Handling complex meta-analytic data structures using robust variance estimates: A tutorial in R. Journal of Developmental and Life-Course Criminology. 2016; 2; 1: 85-112. 10.1007/s40865-016-0026-5</bibtext> </blist> <blist> <bibtext> *Tezler, E. G. (1993). The effects of modeled strategies and attributions on students' self-regulated learning and spelling achievement (Publication No. 9325155) [Doctoral dissertation]. The City University of New York.</bibtext> </blist> <blist> <bibtext> *Thibodeau, A. E. (1964). Improving composition writing with grammar and organization exercises utilizing differentiated group patterns [Unpublished doctoral dissertation]. Boston University.</bibtext> </blist> <blist> <bibtext> *Thomas, M-L. (1995). The effect of genre-specific story grammar instruction on recall, comprehension, and writing of tenth-grade English students (Publication No. 9626945) [Doctoral dissertation]. Marquette University.</bibtext> </blist> <blist> <bibtext> Tipton E, Pustejovsky JE. Small-sample adjustments for tests of moderators and model fit using robust variance estimation in meta-regression. Journal of Educational and Behavioral Statistics. 2015; 40; 6: 604-634. 10.3102/1076998615606099</bibtext> </blist> <blist> <bibtext> Tolchinsky, L. (2016). From text to language and back again: The emergence of written language. In C. A. MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (2nd ed, pp. 144–159). Guilford Press.</bibtext> </blist> <blist> <bibtext> Tukey JW. Exploratory data analysis. 1977; Addison Wesley</bibtext> </blist> <blist> <bibtext> Vahidi A, Karimi L, Mahmoodi MH. The effect of reconstruction as a noticing strategy on Iranian female first grade high school students' writing ability. Theory and Practice in Language Studies. 2016; 6; 2: 310-324. 10.17507/tpls.0602.12</bibtext> </blist> <blist> <bibtext> *van Beuningen, C, G, de Jong, N. H, & Kuiken, F. (2012). Evidence on the effectiveness of comprehensive error correction in second language writing. Language Learning, 62(1), 1–41. https://doi.org/10.1111/j.1467-9922.2011.00674.x</bibtext> </blist> <blist> <bibtext> van Drie J, van Boxtel C, Erkens G, Kanselaar G. Using representational tools to support historical reasoning in computer-supported collaborative learning. Technology, Pedagogy and Education. 2005; 14; 1: 25-42. 10.1080/14759390500200191</bibtext> </blist> <blist> <bibtext> van Drie J, van Driel J, van Weijen D. Developing students' writing in History: Effects of a teacher-designed domain-specific writing instruction. Journal of Writing Research. 2021; 13; 2: 201-229. 10.17239/jowr-2021.13.02.01</bibtext> </blist> <blist> <bibtext> van Driel J, van Drie J, van Boxtel C. Writing about historical significance: The effects of a reading-to-write instruction. International Journal of Educational Research. 2022; 122: 1-13. 10.1016/j.ijer.2022.101924</bibtext> </blist> <blist> <bibtext> *van Wagenen, D. A. (1988). Computerized prewriting activities and the writing performance of high school juniors [Unpublished doctoral dissertation]. The George Washington University.</bibtext> </blist> <blist> <bibtext> *Vinson, L. L. N. (1971). The effects of two prewriting activities upon the overall quality of ninth graders' descriptive paragraphs [Unpublished doctoral dissertation] University of South Carolina.</bibtext> </blist> <blist> <bibtext> *Wagner, J. H. (1978). Peer teaching in spelling: An experimental study in selected Seventh-day Adventist high schools (Publication No. 7.913326) [Doctoral dissertation]. University of Florida.</bibtext> </blist> <blist> <bibtext> *Walker, R. R. (1970). A comparison of an individualized and a group-directed system for teaching spelling in the eighth grade [Unpublished doctoral dissertation]. Columbia University.</bibtext> </blist> <blist> <bibtext> *Widvey, L. I. H. (1971). A study of the use of a problem-solving approach to composition in high school English [Unpublished doctoral dissertation]. The University of Nebraska-Lincoln.</bibtext> </blist> <blist> <bibtext> Wilson J, Czik A. Automated essay evaluation software in English Language Arts classrooms: Effects on teacher feedback, student motivation, and writing quality. Computers & Education. 2016; 100: 94-109. 10.1016/j.compedu.2016.05.004</bibtext> </blist> <blist> <bibtext> Wilson J, Roscoe RD. Automated writing evaluation and feedback: Multiple metrics of efficacy. Journal of Educational Computing Research. 2019; 58; 1: 87-125. 10.1177/0735633119830764</bibtext> </blist> <blist> <bibtext> *Wise, W. G, & Slater, W. H. (1992). The effects of revision instruction on eighth graders' persuasive writing (Publication No. 9304422) [Doctoral dissertation]. University of Maryland College Park.</bibtext> </blist> <blist> <bibtext> Wissinger DR, de la Paz S. Effects of critical discussions on middle school students' written historical arguments. Journal of Educational Psychology. 2016; 108; 1: 43-59. 10.1037/edu0000043</bibtext> </blist> <blist> <bibtext> Wissinger DR, De La Paz S, Jackson C. The effects of historical reading and writing strategy instruction with fourth-through sixth-grade students. Journal of Educational Psychology. 2021; 113; 1: 49-67. 10.1037/edu0000463</bibtext> </blist> <blist> <bibtext> Wong BYL, Hoskyn M, Jai D, Ellis P, Watson K. The comparative efficacy of two approaches to teaching sixth graders opinion essay writing. Contemporary Educational Psychology. 2008; 33; 4: 757-784. 10.1016/j.cedpsych.2007.12.004</bibtext> </blist> <blist> <bibtext> Yeh SS. Empowering education: Teaching argumentative writing to cultural minority middle-school students. Research in the Teaching of English. 1998; 33; 1: 49-83. 10.58680/rte19983912</bibtext> </blist> <blist> <bibtext> Zellermayer M, Salomon G, Globerson T, Givon H. Enhancing writing-related metacognitions through a computerized writing partner. American Educational Research Journal. 1991; 28; 2: 373-391. 10.3102/00028312028002373</bibtext> </blist> </ref> <aug> <p>By Steve Graham; Yucheng Cao; Young-Suk Grace Kim; Joongwon Lee; Tamara Tate; Penelope Collins; Minkyung Cho; Youngsun Moon; Huy Quoc Chung and Carol Booth Olson</p> <p>Reported by Author; Author; Author; Author; Author; Author; Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib62" firstref="ref2"></nolink> <nolink nlid="nl2" bibid="bib65" firstref="ref3"></nolink> <nolink nlid="nl3" bibid="bib56" firstref="ref4"></nolink> <nolink nlid="nl4" bibid="bib57" firstref="ref6"></nolink> <nolink nlid="nl5" bibid="bib156" firstref="ref7"></nolink> <nolink nlid="nl6" bibid="bib77" firstref="ref8"></nolink> <nolink nlid="nl7" bibid="bib118" firstref="ref9"></nolink> <nolink nlid="nl8" bibid="bib68" firstref="ref10"></nolink> <nolink nlid="nl9" bibid="bib41" firstref="ref12"></nolink> <nolink nlid="nl10" bibid="bib105" firstref="ref14"></nolink> <nolink nlid="nl11" bibid="bib66" firstref="ref15"></nolink> <nolink nlid="nl12" bibid="bib98" firstref="ref16"></nolink> <nolink nlid="nl13" bibid="bib53" firstref="ref18"></nolink> <nolink nlid="nl14" bibid="bib59" firstref="ref19"></nolink> <nolink nlid="nl15" bibid="bib63" firstref="ref20"></nolink> <nolink nlid="nl16" bibid="bib78" firstref="ref21"></nolink> <nolink nlid="nl17" bibid="bib116" firstref="ref22"></nolink> <nolink nlid="nl18" bibid="bib138" firstref="ref23"></nolink> <nolink nlid="nl19" bibid="bib67" firstref="ref24"></nolink> <nolink nlid="nl20" bibid="bib144" firstref="ref25"></nolink> <nolink nlid="nl21" bibid="bib64" firstref="ref26"></nolink> <nolink nlid="nl22" bibid="bib87" firstref="ref29"></nolink> <nolink nlid="nl23" bibid="bib147" firstref="ref30"></nolink> <nolink nlid="nl24" bibid="bib17" firstref="ref32"></nolink> <nolink nlid="nl25" bibid="bib145" firstref="ref34"></nolink> <nolink nlid="nl26" bibid="bib120" firstref="ref35"></nolink> <nolink nlid="nl27" bibid="bib61" firstref="ref38"></nolink> <nolink nlid="nl28" bibid="bib12" firstref="ref53"></nolink> <nolink nlid="nl29" bibid="bib150" firstref="ref54"></nolink> <nolink nlid="nl30" bibid="bib130" firstref="ref55"></nolink> <nolink nlid="nl31" bibid="bib71" firstref="ref60"></nolink> <nolink nlid="nl32" bibid="bib42" firstref="ref78"></nolink> <nolink nlid="nl33" bibid="bib24" firstref="ref82"></nolink> <nolink nlid="nl34" bibid="bib157" firstref="ref94"></nolink> <nolink nlid="nl35" bibid="bib137" firstref="ref95"></nolink> <nolink nlid="nl36" bibid="bib100" firstref="ref96"></nolink> <nolink nlid="nl37" bibid="bib25" firstref="ref97"></nolink> <nolink nlid="nl38" bibid="bib72" firstref="ref98"></nolink> <nolink nlid="nl39" bibid="bib26" firstref="ref99"></nolink> <nolink nlid="nl40" bibid="bib132" firstref="ref100"></nolink> <nolink nlid="nl41" bibid="bib143" firstref="ref101"></nolink> <nolink nlid="nl42" bibid="bib47" firstref="ref102"></nolink> <nolink nlid="nl43" bibid="bib151" firstref="ref103"></nolink> <nolink nlid="nl44" bibid="bib73" firstref="ref104"></nolink> <nolink nlid="nl45" bibid="bib155" firstref="ref105"></nolink> <nolink nlid="nl46" bibid="bib81" firstref="ref106"></nolink> <nolink nlid="nl47" bibid="bib174" firstref="ref108"></nolink> <nolink nlid="nl48" bibid="bib133" firstref="ref109"></nolink> <nolink nlid="nl49" bibid="bib34" firstref="ref110"></nolink> <nolink nlid="nl50" bibid="bib169" firstref="ref111"></nolink> <nolink nlid="nl51" bibid="bib101" firstref="ref112"></nolink> <nolink nlid="nl52" bibid="bib31" firstref="ref113"></nolink> <nolink nlid="nl53" bibid="bib93" firstref="ref114"></nolink> <nolink nlid="nl54" bibid="bib22" firstref="ref115"></nolink> <nolink nlid="nl55" bibid="bib162" firstref="ref116"></nolink> <nolink nlid="nl56" bibid="bib39" firstref="ref117"></nolink> <nolink nlid="nl57" bibid="bib129" firstref="ref118"></nolink> <nolink nlid="nl58" bibid="bib85" firstref="ref119"></nolink> <nolink nlid="nl59" bibid="bib119" firstref="ref120"></nolink> <nolink nlid="nl60" bibid="bib103" firstref="ref121"></nolink> <nolink nlid="nl61" bibid="bib60" firstref="ref154"></nolink> <nolink nlid="nl62" bibid="bib58" firstref="ref158"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1468906
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Effective Writing Instruction for Students in Grades 6 to 12: A Best Evidence Meta-Analysis
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Steve+Graham%22">Steve Graham</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-6702-5865">0000-0002-6702-5865</externalLink>)<br /><searchLink fieldCode="AR" term="%22Yucheng+Cao%22">Yucheng Cao</searchLink><br /><searchLink fieldCode="AR" term="%22Young-Suk+Grace+Kim%22">Young-Suk Grace Kim</searchLink><br /><searchLink fieldCode="AR" term="%22Joongwon+Lee%22">Joongwon Lee</searchLink><br /><searchLink fieldCode="AR" term="%22Tamara+Tate%22">Tamara Tate</searchLink><br /><searchLink fieldCode="AR" term="%22Penelope+Collins%22">Penelope Collins</searchLink><br /><searchLink fieldCode="AR" term="%22Minkyung+Cho%22">Minkyung Cho</searchLink><br /><searchLink fieldCode="AR" term="%22Youngsun+Moon%22">Youngsun Moon</searchLink><br /><searchLink fieldCode="AR" term="%22Huy+Quoc+Chung%22">Huy Quoc Chung</searchLink><br /><searchLink fieldCode="AR" term="%22Carol+Booth+Olson%22">Carol Booth Olson</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Reading+and+Writing%3A+An+Interdisciplinary+Journal%22"><i>Reading and Writing: An Interdisciplinary Journal</i></searchLink>. 2025 38(4):1-46.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Springer. Available from: Springer Nature. One New York Plaza, Suite 4600, New York, NY 10004. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-460-1700; e-mail: customerservice@springernature.com; Web site: https://link.springer.com/
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 46
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: SourceSuprt
  Label: Sponsoring Agency
  Group: SrcSuprt
  Data: Institute of Education Sciences (ED)
– Name: NumberContract
  Label: Contract Number
  Group: NumCntrct
  Data: R305C190007
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Information Analyses<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Secondary+Education%22">Secondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Writing+Instruction%22">Writing Instruction</searchLink><br /><searchLink fieldCode="DE" term="%22Secondary+Education%22">Secondary Education</searchLink><br /><searchLink fieldCode="DE" term="%22Instructional+Effectiveness%22">Instructional Effectiveness</searchLink><br /><searchLink fieldCode="DE" term="%22Writing+Skills%22">Writing Skills</searchLink><br /><searchLink fieldCode="DE" term="%22Educational+Research%22">Educational Research</searchLink><br /><searchLink fieldCode="DE" term="%22Research+Methodology%22">Research Methodology</searchLink><br /><searchLink fieldCode="DE" term="%22Meta+Analysis%22">Meta Analysis</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1007/s11145-024-10539-2
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0922-4777<br />1573-0905
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: The current best evidence meta-analysis reanalyzed the data from a meta-analysis by Graham et al. (J Educ Psychol 115:1004-1027, 2023). This meta-analysis and the prior one examined if teaching writing improved the writing of students in Grades 6 to 12, examining effects from writing intervention studies employing experimental and quasi-experimental designs (with pretests). In contrast to the prior meta-analysis, we eliminated all N of 1 treatment/control comparisons, studies with an attrition rate over 20%, studies that did not control for teacher effects, and studies that did not contain at least one reliable writing measure (0.70 or greater). Any writing outcome that was not reliable was also eliminated. Across 148 independent treatment/control comparisons, yielding 1,076 writing effect sizes (ESs) involving 22,838 students, teaching writing resulted in a positive and statistically detectable impact on students' writing (ES = 0.38). Further, six of the 10 writing treatments tested in four or more independent comparisons improved students' performance. This included the process approach to writing (0.75), strategy instruction (0.59), transcription instruction (0.54), feedback (0.30), pre-writing activities (0.32), and peer assistance (0.59). In addition, the Self-Regulated Strategy Development model for teaching writing strategies yielded a statistically significant ES of 0.84, whereas other approaches to teaching writing strategies resulted in a statistically significant ES of 0.51. The findings from this meta-analysis and the Graham et al. (2023) review which included studies that were methodologically weaker were compared. Implications for practice, research, and theory are presented.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: CodeSource
  Label: IES Funded
  Group: SrcInfo
  Data: Yes
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2025
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1468906
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1468906
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1007/s11145-024-10539-2
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 46
        StartPage: 1
    Subjects:
      – SubjectFull: Writing Instruction
        Type: general
      – SubjectFull: Secondary Education
        Type: general
      – SubjectFull: Instructional Effectiveness
        Type: general
      – SubjectFull: Writing Skills
        Type: general
      – SubjectFull: Educational Research
        Type: general
      – SubjectFull: Research Methodology
        Type: general
      – SubjectFull: Meta Analysis
        Type: general
    Titles:
      – TitleFull: Effective Writing Instruction for Students in Grades 6 to 12: A Best Evidence Meta-Analysis
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Steve Graham
      – PersonEntity:
          Name:
            NameFull: Yucheng Cao
      – PersonEntity:
          Name:
            NameFull: Young-Suk Grace Kim
      – PersonEntity:
          Name:
            NameFull: Joongwon Lee
      – PersonEntity:
          Name:
            NameFull: Tamara Tate
      – PersonEntity:
          Name:
            NameFull: Penelope Collins
      – PersonEntity:
          Name:
            NameFull: Minkyung Cho
      – PersonEntity:
          Name:
            NameFull: Youngsun Moon
      – PersonEntity:
          Name:
            NameFull: Huy Quoc Chung
      – PersonEntity:
          Name:
            NameFull: Carol Booth Olson
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 04
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 0922-4777
            – Type: issn-electronic
              Value: 1573-0905
          Numbering:
            – Type: volume
              Value: 38
            – Type: issue
              Value: 4
          Titles:
            – TitleFull: Reading and Writing: An Interdisciplinary Journal
              Type: main
ResultId 1