View in EDS HTML Full Text PDF Full Text

What Drives Student Engagement and Learning in Video Lectures? An Investigation of Instructor Visibility, Playback Speed, and Student Preferences

Saved in:
Bibliographic Details
Title: What Drives Student Engagement and Learning in Video Lectures? An Investigation of Instructor Visibility, Playback Speed, and Student Preferences
Language: English
Authors: Dahwi Ahn (ORCID 0000-0002-4300-6520), Jason C. K. Chan
Source: Applied Cognitive Psychology. 2025 39(2).
Availability: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
Peer Reviewed: Y
Page Count: 18
Publication Date: 2025
Sponsoring Agency: National Science Foundation (NSF)
Contract Number: 2017333
Document Type: Journal Articles
Reports - Research
Tests/Questionnaires
Descriptors: Learner Engagement, Video Technology, Lecture Method, Student Attitudes, Preferences, Educational Technology, Time Factors (Learning), Learning Processes, Electronic Learning
DOI: 10.1002/acp.70026
ISSN: 0888-4080
1099-0720
Abstract: COVID-19 greatly increased the online delivery of higher education. But one limitation of online learning is that students often struggle to stay engaged while watching online lectures. We examined whether including an instructor's face in lecture videos (instructor visibility) enhances student engagement or learning. In two preregistered experiments, we found that instructor visibility in lecture videos did not affect either engagement or learning overall. However, participants reported higher engagement when they watched a video that aligned with their preference for instructor visibility. For example, participants who favored videos with the instructor visible reported greater engagement with such videos compared to those without the instructor, and vice versa. Additionally, we examined the effects of playback speed on engagement and learning. Our results suggest that speeded playing did not impact engagement but resulted in better learning efficiency. Lastly, using GPT, we explored participants' open-ended responses to understand their preference for video lectures.
Abstractor: As Provided
Entry Date: 2025
Accession Number: EJ1468318
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwFNJamo7kXurSR0suNINKQWAAAA4jCB3wYJKoZIhvcNAQcGoIHRMIHOAgEAMIHIBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDCgNT-QIm8899hRy9QIBEICBmr5hpAz3zPeBdIz9feLRE_q_7MS4JU-GCK_fsrxkJxe4yrbLmtmz7WvqdGSqPuGbMlW_uDDK-JLgypg66gtf0oERD_c-jH-YlrGFwdFf897Qo2WlcWm4ZrCf5d1IZCqlTd9p-UEmvpBqCz7vFpILwWfyBgKj_OaeFHL1eOJA5LaOttOLu2briRrl1C8lwe4p4vK65Wwn5n1faRM=
Text:
  Availability: 1
  Value: <anid>AN0184623579;bu801mar.25;2025Apr23.01:41;v2.2.500</anid> <title id="AN0184623579-1">What Drives Student Engagement and Learning in Video Lectures? An Investigation of Instructor Visibility, Playback Speed, and Student Preferences </title> <p>COVID‐19 greatly increased the online delivery of higher education. But one limitation of online learning is that students often struggle to stay engaged while watching online lectures. We examined whether including an instructor's face in lecture videos (instructor visibility) enhances student engagement or learning. In two preregistered experiments, we found that instructor visibility in lecture videos did not affect either engagement or learning overall. However, participants reported higher engagement when they watched a video that aligned with their preference for instructor visibility. For example, participants who favored videos with the instructor visible reported greater engagement with such videos compared to those without the instructor, and vice versa. Additionally, we examined the effects of playback speed on engagement and learning. Our results suggest that speeded playing did not impact engagement but resulted in better learning efficiency. Lastly, using GPT, we explored participants' open‐ended responses to understand their preference for video lectures.</p> <p>Keywords: engagement; individual preference; instructor presence; instructor visibility; online learning; video playback speed</p> <hd id="AN0184623579-2">Introduction</hd> <p>The transition of higher education to online delivery has been accelerated since COVID‐19 (Cameron et al. [<reflink idref="bib8" id="ref1">8</reflink>]; Chan and Ahn [<reflink idref="bib12" id="ref2">12</reflink>]). For example, in fall 2022, nearly half of the undergraduates (i.e., 10 million students) in the United States were enrolled in online courses (National Center for Education Statistics [<reflink idref="bib56" id="ref3">56</reflink>]). This trend is also exemplified by platforms like Coursera or MOOCs, which provide online lectures from many universities. To date, Coursera has been used by 148 million people and MOOCs by 35 million (Coursera [<reflink idref="bib16" id="ref4">16</reflink>]; MOOC.org [<reflink idref="bib51" id="ref5">51</reflink>]). A key feature that has facilitated the universal spread of online learning is recorded video lectures, which offer great flexibility and make education more affordable and accessible to a wide audience.</p> <p>While this flexibility provides many benefits, it may negatively impact student engagement as students can easily be distracted while watching lectures at their convenience. For example, students often struggle to maintain attention during online lectures after only a short period (Ashburner et al. [<reflink idref="bib5" id="ref6">5</reflink>]), such as under 10 min (Manasrah, Masoud, and Jaradat [<reflink idref="bib45" id="ref7">45</reflink>]; Ozan and Ozarslan [<reflink idref="bib58" id="ref8">58</reflink>]) or even under 6 min (Guo, Kim, and Rubin [<reflink idref="bib30" id="ref9">30</reflink>]). Furthermore, 50%–70% of students reported not being engaged during online lectures (Ahn [<reflink idref="bib2" id="ref10">2</reflink>]; Ahn et al. [<reflink idref="bib4" id="ref11">4</reflink>]; Was, Hollis, and Dunlosky [<reflink idref="bib92" id="ref12">92</reflink>]). However, maintaining student engagement is crucial because it is associated with positive learning outcomes such as higher GPA (Kuh [<reflink idref="bib38" id="ref13">38</reflink>], [<reflink idref="bib39" id="ref14">39</reflink>]; Kuh et al. [<reflink idref="bib40" id="ref15">40</reflink>]), retention (Aguiar et al. [<reflink idref="bib1" id="ref16">1</reflink>]; Gray and Perkins [<reflink idref="bib28" id="ref17">28</reflink>]; Matz et al. [<reflink idref="bib47" id="ref18">47</reflink>]), and future enrollment (Finn [<reflink idref="bib25" id="ref19">25</reflink>]; Fraysier, Reschly, and Appleton [<reflink idref="bib26" id="ref20">26</reflink>]). Thus, in this study, we aimed to investigate ways to improve engagement in online lectures.</p> <p>One way to increase students' engagement is by enhancing <emph>instructor presence</emph>, which makes students feel that the instructor is actively involved (Dixson [<reflink idref="bib20" id="ref21">20</reflink>]). Instructor presence includes various aspects, such as being available and responsive to students' needs, fostering a sense of community, and making the class feel "live" (Richardson et al. [<reflink idref="bib66" id="ref22">66</reflink>]; Sheridan and Kelly [<reflink idref="bib72" id="ref23">72</reflink>]). In the context of recorded lectures, the simplest way to ensure instructor presence is by including instructor's face in the video, which can be defined as <emph>instructor visibility</emph>. In the current study, we specifically examined the effect of instructor visibility in video lectures on student engagement and learning. For exposition purposes, instructor‐visible refers to a lecture video in which participants can see the face of the instructor, and instructor‐not‐visible refers to a lecture in which the instructor is not shown on screen. Note that in both situations, participants can hear the instructor speak.</p> <p>Previous studies have suggested that instructor‐visible videos have benefits on student engagement, such as higher satisfaction and positive affect, compared to instructor‐not‐visible videos (Kizilcec, Papadopoulos, and Sritanyaratana [<reflink idref="bib36" id="ref24">36</reflink>]; Li et al. [<reflink idref="bib44" id="ref25">44</reflink>]; Sondermann and Merkt [<reflink idref="bib77" id="ref26">77</reflink>]; Wilson et al. [<reflink idref="bib94" id="ref27">94</reflink>]; Wang and Antonenko [<reflink idref="bib90" id="ref28">90</reflink>]; Wong et al. [<reflink idref="bib95" id="ref29">95</reflink>]). However, few studies have examined whether individual preference for instructor visibility might influence learning‐related outcomes (cf. Kizilcec, Bailenson, and Gomez [<reflink idref="bib35" id="ref30">35</reflink>]). For example, would instructor‐visible videos (vs. instructor‐not‐visible videos) induce higher engagement only for students who prefer to see instructors in lectures, but not for students who prefer the opposite? A novel contribution of the present study is that we view engagement through the lens of individual preference. In Experiments 1 and 2, we investigated whether the effect of instructor visibility on engagement interacts with individual preference. Additionally, we explored the reasons behind individual preference for instructor visibility.</p> <p>Another way to potentially increase engagement is by increasing the <emph>playback speed</emph> of lecture videos, which can save time and might increase the efficiency of learning (Ashburner et al. [<reflink idref="bib5" id="ref31">5</reflink>]; Murphy et al. [<reflink idref="bib53" id="ref32">53</reflink>]; Risko, Liu, and Bianchi [<reflink idref="bib67" id="ref33">67</reflink>]). In Experiment 2, we examined the effect of video playback speed on engagement and whether this effect interacted with instructor visibility. Lastly, in both Experiments 1 and 2, we also investigated the effects of instructor visibility and playback speed on <emph>learning</emph>.</p> <p>Below, we briefly review previous studies regarding the effects of instructor visibility on engagement and learning, and how it may interact with individual preference.</p> <hd id="AN0184623579-3">The Effect of Instructor Visibility on Engagement</hd> <p>Broadly speaking, engagement is defined as one's time and effort dedicated to educational activities (Krause and Coates [<reflink idref="bib37" id="ref34">37</reflink>]; Kuh [<reflink idref="bib39" id="ref35">39</reflink>]), and it has been investigated in different ways depending on the context and the level at which it is conceptualized (Sinatra, Heddy, and Lombardi [<reflink idref="bib75" id="ref36">75</reflink>]). At a macro level, research on engagement focuses on overall student behaviors across or within courses, semesters, or universities, whereas at a micro level, research focuses on a student's engagement on a momentary or task basis. Engagement can be measured through various methods, including experience sampling, eye‐tracking, and observations, but the most common method is self‐report (Fredricks [<reflink idref="bib27" id="ref37">27</reflink>]).</p> <p>The current study focused on micro‐level engagement using self‐report measures. Specifically, we examined emotional and behavioral engagement (Sinatra, Heddy, and Lombardi [<reflink idref="bib75" id="ref38">75</reflink>]; Wang et al. [<reflink idref="bib91" id="ref39">91</reflink>]). Emotional engagement is characterized as the manifestation of positive emotional responses toward learning activities. Although previous studies on the effect of instructor visibility did not always use this term, emotional engagement aligns with metrics used in prior studies such as affect, liking, and satisfaction (Kizilcec, Papadopoulos, and Sritanyaratana [<reflink idref="bib36" id="ref40">36</reflink>]; Wang and Antonenko [<reflink idref="bib90" id="ref41">90</reflink>]; Wilson et al. [<reflink idref="bib94" id="ref42">94</reflink>]). Several studies showed that presenting an instructor's face in lecture videos can increase emotional engagement. For example, Wang and Antonenko ([<reflink idref="bib90" id="ref43">90</reflink>]) found that students reported greater satisfaction and believed they learned better when they could see the instructor's face during the lecture. Similarly, learners perceived an instructor‐visible lecture as more enjoyable, helpful, and interesting than an instructor‐not‐visible lecture (Kizilcec, Papadopoulos, and Sritanyaratana [<reflink idref="bib36" id="ref44">36</reflink>]; Wilson et al. [<reflink idref="bib94" id="ref45">94</reflink>]).</p> <p>Behavioral engagement is the degree to which a student is actively involved in their learning. According to Wong et al. ([<reflink idref="bib95" id="ref46">95</reflink>]), presenting the instructor's face in a lecture video can give students a sense of connection with the instructor and encourage positive in‐class behaviors. In their study, participants watched either a talking‐head video of the instructor, a lecture hall recording with the instructor relatively far away, or just saw the slides with voice narration. Compared with lecture hall recording or voice narration, participants rated the talking‐head video higher in instructor likability, expertise, and immediacy (thus showing increased emotional engagement). Further, participants indicated that they would feel more connected to the instructor and would be more willing to participate in class if they saw the instructor's face up close (rather than in a lecture hall recording), thus showing increased behavioral engagement.</p> <hd id="AN0184623579-4">The Effect of Instructor Visibility on Learning</hd> <p>Learning can be assessed in various ways. For example, a memory test examines whether individuals remember the content of a lecture, whereas a comprehension (or transfer) test goes beyond memorization and requires individuals to apply what they have learned in a different context. In the current study, we define learning as performance in either test format.</p> <p>Empirical findings about the effect of instructor visibility on learning are mixed, with some studies reporting advantages, others reporting disadvantages, and some showing null impact. According to the personalization principle (Clark and Mayer [<reflink idref="bib15" id="ref47">15</reflink>]; Mayer [<reflink idref="bib48" id="ref48">48</reflink>]), the instructor's face in a lecture video might provide social cues and elicit the type of social responses from learners during in‐person lectures, thereby encouraging deeper cognitive processing (Reeves and Nass [<reflink idref="bib65" id="ref49">65</reflink>]). Supporting this idea, some studies have shown that students learn better from a pedagogical agent, a conversational style humanlike character that guides learners, than from a formal text (Davis [<reflink idref="bib19" id="ref50">19</reflink>]; Moreno et al. [<reflink idref="bib52" id="ref51">52</reflink>]).</p> <p>Instructor‐visible videos can also facilitate online learning by visually guiding learners to where they should pay attention (Li et al. [<reflink idref="bib44" id="ref52">44</reflink>]). For example, an instructor can point out the bullet point that she was explaining or use gestures to demonstrate movements. These attention‐directing behaviors might be especially beneficial for learning complex knowledge. For example, van Gog, Verveer, and Verveer ([<reflink idref="bib87" id="ref53">87</reflink>]) had participants watch two presentations of a video on how to solve a complex puzzle (similar to the Tower of Hanoi).[<reflink idref="bib1" id="ref54">1</reflink>] In both viewings, the instructor's face was either visible or not. After watching each video, participants were asked to solve a novel but similar puzzle. Critically, participants who watched the video with the instructor's face performed better on the puzzle after the second viewing. Van Gog and colleagues attributed this advantage to participants using the instructor's eye gaze to anticipate the next step during the second viewing (cf. van Wermeskerken and van Gog [<reflink idref="bib88" id="ref55">88</reflink>]).</p> <p>However, some studies have also suggested that including an instructor's face could harm learning. This is because the instructor's face can impart cognitive load that is unnecessary for understanding the content (Harp and Mayer [<reflink idref="bib31" id="ref56">31</reflink>]; Sweller [<reflink idref="bib81" id="ref57">81</reflink>]; Wang and Antonenko [<reflink idref="bib90" id="ref58">90</reflink>]). Information presentation is transient in video lectures, so learners must pay timely attention. Focusing on the instructor's face might in turn reduce one's attention to the lecture content and undermine learning (Sepp et al. [<reflink idref="bib71" id="ref59">71</reflink>]; van Wermeskerken and van Gog [<reflink idref="bib88" id="ref60">88</reflink>]).</p> <p>Consistent with this expectation, Wilson et al. ([<reflink idref="bib94" id="ref61">94</reflink>]) reported that instructor visibility in lecture videos impaired student comprehension. Participants watched either a lecture hall video with the instructor present or a static image of an empty lecture hall with audio narration. Those who listened to the audio narration outperformed their lecture hall counterparts on a comprehension test. When explaining this result, Wilson and colleagues suggested that the instructor's face served as a seductive detail that captured the learner's attention without contributing to actual learning. Consistent with this idea, Marcus, King, and Markant ([<reflink idref="bib46" id="ref62">46</reflink>]) also found that the duration that students spent looking at the instructor's face was not related to learning.</p> <p>These mixed findings suggest that the effect of instructor visibility on learning might depend on various factors. These factors may include the type of material that students are attempting to learn and the extent of attention‐directing behaviors exhibited by the instructor. For example, instructor visibility might be more beneficial for learning materials in which paying timely attention to visuals is important, but only if instructors provide helpful attention‐directing behaviors. However, examining when instructor visibility helps or harms learning was not the goal of the current study. Without these kinds of moderators being specified, the coexistence of advantages and disadvantages in instructor visibility on learning may lead to a neutral effect, as reported in several studies (Kizilcec, Papadopoulos, and Sritanyaratana [<reflink idref="bib36" id="ref63">36</reflink>]; Mayer, Dow, and Mayer [<reflink idref="bib49" id="ref64">49</reflink>]; Moreno et al. [<reflink idref="bib52" id="ref65">52</reflink>]; van Wermeskerken and van Gog [<reflink idref="bib88" id="ref66">88</reflink>]).</p> <hd id="AN0184623579-5">Individual Preference and Instructor Visibility</hd> <p>Although multiple experimental studies have examined the influence of instructor visibility on engagement and learning in lecture videos, little is known about whether this effect depends on individual preference (cf. Kizilcec, Bailenson, and Gomez [<reflink idref="bib35" id="ref67">35</reflink>]). What people find engaging is subjective and can vary across individuals (Groccia [<reflink idref="bib29" id="ref68">29</reflink>]; Park, Holloway, and Arendtsz [<reflink idref="bib59" id="ref69">59</reflink>]; Taylor and Statler [<reflink idref="bib84" id="ref70">84</reflink>]), so some students may prefer to see the instructor whereas others feel the opposite. Would the former student be more engaged and learn better after watching an instructor‐visible lecture and vice‐versa? What about students who have no preference?</p> <p>In a representative sample study (<emph>N</emph> = 19,088), Kizilcec, Bailenson, and Gomez ([<reflink idref="bib35" id="ref71">35</reflink>]) examined the effect of instructor preference on affect and perceived learning, both of which are aligned with emotional engagement.[<reflink idref="bib2" id="ref72">2</reflink>] Specifically, the researchers gave students the option to watch lecture videos either with or without the instructor and found that a majority of students (57%) chose to watch the instructor‐visible videos. However, a considerable minority (35%) chose the instructor‐not‐visible videos, indicating that students had different preferences regarding instructor visibility. Furthermore, students who selected the instructor‐visible videos rated their learning experience more positively than those who chose the instructor‐not‐visible videos. However, in Kizilcec et al., because students always watched the videos they preferred, it remains unclear what would occur if students' preferences were dishonored. In real life, students' preference may not be honored because they must watch whatever lecture videos are provided by the instructor. To our knowledge, no studies have investigated whether individual preference interacts with instructor visibility. This is an important question; if the effects of instructor visibility depend on an individual's preference, instructors might optimize student engagement by providing two different versions of video lectures.</p> <p>In contrast to the idea that individual preference might matter for engagement, it is less likely that learning would be affected by individual preference given findings from the learning style literature. Learning style refers to the concept that people believe certain instructional styles work best for them. For example, some learners self‐identify as visual learners, claiming that they can grasp concepts better with visual materials, whereas others consider themselves auditory learners, asserting that they learn more effectively with auditory information. However, Pashler et al. ([<reflink idref="bib60" id="ref73">60</reflink>]) argued that scant empirical evidence supports the idea that individuals learn best when they are taught with their preferred style. Subsequent studies have largely confirmed that learning styles do not significantly impact learning (Cuevas [<reflink idref="bib17" id="ref74">17</reflink>]; Fallace [<reflink idref="bib23" id="ref75">23</reflink>]; Husmann and O'Loughlin [<reflink idref="bib33" id="ref76">33</reflink>]; Rogowsky, Calhoun, and Tallal [<reflink idref="bib68" id="ref77">68</reflink>], [<reflink idref="bib69" id="ref78">69</reflink>]). Both concepts—individual preference toward instructor visibility and learning styles—revolve around the idea of personal preference, so it is possible that individual preference for instructor visibility would not influence learning.</p> <p>We must note here that even if individual preference does not affect learning from instructor visible/not‐visible lectures, it does not mean that individual preference is unimportant, because it can still affect engagement. As suggested earlier, student engagement itself can have a positive long‐term influence, such as whether students continue to study in the same field, what major they choose, or if they would re‐enroll in the future (Aguiar et al. [<reflink idref="bib1" id="ref79">1</reflink>]; Finn [<reflink idref="bib25" id="ref80">25</reflink>]; Fraysier, Reschly, and Appleton [<reflink idref="bib26" id="ref81">26</reflink>]; Gray and Perkins [<reflink idref="bib28" id="ref82">28</reflink>]; Matz et al. [<reflink idref="bib47" id="ref83">47</reflink>]).</p> <hd id="AN0184623579-6">The Current Study</hd> <p>In this study, we investigated the extent to which instructor visibility affects student engagement and learning in a within‐subject design (instructor‐visible vs. instructor‐not‐visible). We predicted that participants would find the instructor‐visible lecture more engaging than the instructor‐not‐visible lecture, given that a majority of previous studies found a beneficial effect of instructor visibility on engagement (Kizilcec, Papadopoulos, and Sritanyaratana [<reflink idref="bib36" id="ref84">36</reflink>]; Wang, Antonenko, and Dawson [<reflink idref="bib89" id="ref85">89</reflink>]; Wong et al. [<reflink idref="bib95" id="ref86">95</reflink>], cf. Sondermann and Merkt [<reflink idref="bib78" id="ref87">78</reflink>]). For learning, we predicted a null effect of instructor visibility because the literature is largely split on whether instructor visibility is positive (Clark and Mayer [<reflink idref="bib15" id="ref88">15</reflink>]; van Gog, Verveer, and Verveer [<reflink idref="bib87" id="ref89">87</reflink>]), negative (Sweller [<reflink idref="bib81" id="ref90">81</reflink>]; Wilson et al. [<reflink idref="bib94" id="ref91">94</reflink>]), or has no effects (Kizilcec, Papadopoulos, and Sritanyaratana [<reflink idref="bib36" id="ref92">36</reflink>]; Mayer, Dow, and Mayer [<reflink idref="bib49" id="ref93">49</reflink>]). Lastly and perhaps most importantly, we examined whether the influence of instructor visibility on engagement and learning would depend on individual preference. Given the novelty of this question, we did not have a directional hypothesis. We also explored participants' reason(s) for their individual preferences regarding instructor visibility.</p> <p>One limitation of existing studies is that most used only one lecture video featuring a single instructor (Kizilcec, Papadopoulos, and Sritanyaratana [<reflink idref="bib36" id="ref94">36</reflink>]; van Gog, Verveer, and Verveer [<reflink idref="bib87" id="ref95">87</reflink>]; Wang, Antonenko, and Dawson [<reflink idref="bib89" id="ref96">89</reflink>]; Wong et al. [<reflink idref="bib95" id="ref97">95</reflink>]). Adequate stimulus sampling (Wells and Windschitl [<reflink idref="bib93" id="ref98">93</reflink>]) is particularly important when examining variables related to engagement and learning, given that variations in how an instructor delivers a lecture can have a profound impact on students' engagement and perceived learning (Carpenter et al. [<reflink idref="bib9" id="ref99">9</reflink>]; Carpenter, Witherby, and Tauber [<reflink idref="bib10" id="ref100">10</reflink>]). For example, an observed positive effect of instructor visibility on engagement could be attributed to the instructor's physical attractiveness or speaking fluency, rather than the presence of the instructor per se. Furthermore, prior research suggested a potential gender bias related to the instructor's gender in lecture videos (Ceci, Kahn, and Williams [<reflink idref="bib11" id="ref101">11</reflink>]; Doubleday and Lee [<reflink idref="bib21" id="ref102">21</reflink>]). Thus, our study implemented lecture videos including four different STEM topics taught by both a man and woman instructor, for a total of eight unique lecture‐instructor combinations.</p> <hd id="AN0184623579-7">Data Availability</hd> <p>Both experiments were pre‐registered on the Open Science Framework (OSF) at https://osf.io/jd6v3 (Experiment 1) and https://osf.io/rc3pv (Experiment 2) prior to data collection. The data and Supporting Information are available on the project page (https://osf.io/xebwp/?view_only=78995d51ae1a4acda5002c021f0cf599).</p> <hd id="AN0184623579-8">Experiment 1</hd> <p></p> <hd id="AN0184623579-9">Design and Participants</hd> <p>Instructor visibility was manipulated within‐subjects (instructor‐visible vs. instructor‐not‐visible). To determine the effect size for power analysis, we conducted a pilot study (<emph>N</emph> = 58). Participants watched two 5‐min lecture videos on computer science, one with instructor‐visible and the other with instructor‐not‐visible. The order of conditions was counterbalanced. After watching each video, participants completed a self‐report engagement survey. Participants provided quantitatively higher engagement ratings for the instructor‐visible video than the instructor‐not‐visible video, <emph>t</emph>(<reflink idref="bib57" id="ref103">57</reflink>) = 1.82, <emph>p</emph> = 0.074, <emph>d</emph> = 0.24, and we used this as an effect size estimate. We conducted a power analysis using G*Power (Faul et al. [<reflink idref="bib24" id="ref104">24</reflink>]). The goal was to detect an effect size of <emph>d</emph> = 0.24 with 0.05 alpha probability in a two‐tailed within‐subjects <emph>t</emph>‐test. To achieve 80% power, 139 participants were necessary.</p> <p>Participants were 158 undergraduates from a large university in the Midwest who completed the study for course credits. Due to an experimenter error, a 5‐min pilot study video was used for one counterbalance (i.e., computer science, woman instructor) instead of the intended actual 8‐min experiment video. As a result, data from those participants (<emph>N</emph> = 19) were replaced with data from an equivalent number of participants collected using the correct videos. Data from 14 participants were excluded from analyses based on pre‐registered criteria (see our Supporting Information for details). The final sample included data from 144 participants (97 women, 44 men, and 3 non‐binary; mean age = 19.38).</p> <hd id="AN0184623579-10">Materials</hd> <p></p> <hd id="AN0184623579-11">Lecture Videos</hd> <p>To ensure the generalizability of our findings, we created 8‐min lecture videos on four different STEM topics—computer science, ecology, physics, and statistics. Each topic had one version with a man instructor and another with a woman instructor. Figure 1 shows examples of video screenshots. In total, 16 lecture videos (4 topics × 2 genders × 2 conditions) were used in the experiments.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/BU8/01mar25/acp70026-fig-0001.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="acp70026-fig-0001.jpg" title="1 Examples of video types. Each cell shows an image of a different lecture topic. Each topic included a different pair of actors." /> </p> <p></p> <hd id="AN0184623579-13">Engagement</hd> <p>We constructed our engagement survey based on <emph>The Math and Science Engagement Scales</emph> (Wang et al. [<reflink idref="bib91" id="ref105">91</reflink>]). We chose this scale because our materials featured STEM topics, and the scale was developed to capture student engagement in math and science. Because the original scale also includes items irrelevant to our study such as engagement in a classroom setting (e.g., I complete my homework on time), we specifically chose items that are appropriate for one‐time learning, which included four items measuring behavioral engagement and another four measuring emotional engagement (see Appendix A).</p> <p>Participants responded to each item on a 5‐point Likert scale, and each choice was converted to a number for analyses (Strongly disagree = 1, Disagree = 2, Neither disagree nor agree = 3, Agree = 4, Strongly agree = 5; with some items reverse coded). The behavioral and emotional engagement score was calculated separately by averaging four items on each scale.</p> <hd id="AN0184623579-14">Test Questions</hd> <p>For each topic, we constructed 10 multiple‐choice questions with four response options. These factual questions aimed to measure retention. For example, one question in the ecology topic asked participants, "Which continent are most toucans from?" and the options presented were "North America, South America (answer), Africa, Asia." Example questions from entire topics are provided in Appendix B. The order of questions was randomized for each participant, but the answer options were not randomized.</p> <hd id="AN0184623579-15">Preference Survey</hd> <p>Participants were asked about their preference for instructor visibility both in general situations and in this particular experiment.</p> <p>General preference question:</p> <p>When watching online video lectures, do you prefer to have the speaker appear in the video or do you prefer not to see the speaker? Note that this question refers to your general preference, NOT about the lectures you watched in this experiment per se.</p> <p>Experiment preference question:</p> <p>In this experiment, you have watched the two lectures either with or without the instructor's face. Which one did you prefer?</p> <p>Participants were asked the general preference question before the experimental preference question, and they could select among <emph>one with the instructor's face, one without the instructor's face</emph>, or <emph>no preference</emph>. These options were not randomized. For the experiment preference question, participants were also asked to justify their preference in an open‐ended response (i.e., <emph>Could you tell us why you preferred the lecture with [or without] the instructor's face?</emph>). Participants who chose "<emph>no preference</emph>" were not asked this question.</p> <p>Following the categorical questions, participants answered three quantitative questions on their general preference toward instructor visibility using the same 5‐point scale as the engagement survey. Data for these questions are presented in the Supporting Information.</p> <hd id="AN0184623579-16">Procedure</hd> <p>The experiment was programmed using Qualtrics, and data collection was conducted online. At the beginning of the experiment, participants were told that they would watch two STEM lectures. Participants were also asked to pay attention and learn the lecture content as they would for an actual class.</p> <p>Figure 2 shows the procedure of Experiment 1. In the experiment, participants watched two different lectures, one with the instructor‐visible condition and the other with the instructor‐not‐visible condition. Qualtrics randomly chose two topics for each participant. Half of the participants watched the instructor‐visible lecture first and vice versa. The gender of the instructor was held constant across the instructor‐visible and instructor‐not‐visible videos for each participant. Once participants viewed the first lecture, they completed an engagement survey and then a multiple‐choice test. The same procedure was repeated for the second lecture. Finally, participants completed a survey about their preference for instructor visibility, demographics, and prior knowledge. The prior knowledge question (i.e., Were you familiar with the lecture topic?) was on a 5‐point Likert scale (extremely familiar, very familiar, moderately familiar, slightly familiar, and not familiar at all) to examine any potential effect of prior knowledge on learning or engagement. Participants who chose "extremely familiar" or "very familiar" were considered to have high prior knowledge (<emph>N</emph> = 11). We analyzed the data both with and without these participants, but the patterns did not change, so we included them in the final analysis.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/BU8/01mar25/acp70026-fig-0002.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="acp70026-fig-0002.jpg" title="2 Procedure of Experiment 1. Experiment 2 followed the same procedure as Experiment 1 with two exceptions: (i) the preference survey was administered at the beginning of the experiment, and (ii) the playback speed of the lecture videos was also manipulated." /> </p> <p></p> <hd id="AN0184623579-18">Results</hd> <p>In our analyses, we conducted two‐tailed tests with ⍺ = 0.05 and reported Bayes factors. When a result is significant in null hypothesis testing (NHST), we present <emph>B</emph><subs>10</subs>, where a higher value indicates greater support for the alternative hypothesis relative to the null. In contrast, when a result is not significant, we present <emph>B</emph><subs>01</subs>, where a higher value indicates greater support for the null hypothesis relative to the alternative. We performed all Bayesian analyses with the default priors in JASP (JASP Team [<reflink idref="bib34" id="ref106">34</reflink>]), which was a point‐null for the null hypothesis and a two‐sided Cauchy distribution with a scale factor of 0.707 for the alternative hypothesis.</p> <p>We examined whether the influence of instructor visibility on engagement depended on lecture topic and instructor gender. To this end, we conducted an analysis of variance (ANOVA) with lecture topic, instructor gender, and instructor visibility condition as independent variables and engagement as a dependent variable. None of the interactions was significant, <emph>F</emph>s < 1.07, <emph>p</emph>s > 0.213, <emph>η</emph><subs>p</subs><sups>2</sups> s < 0.02, <emph>B</emph><subs>01</subs>s > 3.14, which indicates that the effect of instructor visibility on engagement did not vary across topic or instructor gender. The same pattern was found for learning. Thus, we conducted the following analyses after collapsing the data across lecture topics and instructor genders.</p> <hd id="AN0184623579-19">Effects of Instructor Visibility on Engagement and Learning</hd> <p>Before conducting the main analyses, we examined the reliability of the engagement survey. Cronbach's <emph>α</emph> was 0.86 for emotional engagement and 0.79 for behavioral engagement, both of which indicate good reliability (Taber [<reflink idref="bib82" id="ref107">82</reflink>]; Tavakol and Dennick [<reflink idref="bib83" id="ref108">83</reflink>]). We also examined the correlation between engagement scores and test performance. A positive correlation was observed in both instructor‐visible condition (<emph>r</emph><subs>emotional</subs>[<reflink idref="bib142" id="ref109">142</reflink>] = 0.33, <emph>p</emph> < 0.001, <emph>B</emph><subs>10</subs> = 254.44; <emph>r</emph><subs>behavioral</subs>[<reflink idref="bib142" id="ref110">142</reflink>] = 0.35, <emph>p</emph> < 0.001, <emph>B</emph><subs><emph>10</emph></subs> = 945.14), and instructor‐not‐visible condition (<emph>r</emph><subs>emotional</subs>[<reflink idref="bib142" id="ref111">142</reflink>] = 0.24, <emph>p</emph> = 0.004, <emph>B</emph><subs><emph>10</emph></subs> = 6.55; <emph>r</emph><subs>behavioral</subs>[<reflink idref="bib142" id="ref112">142</reflink>] = 0.22, <emph>p</emph> = 0.010, <emph>B</emph><subs>10</subs> = 2.86). These results suggest that higher engagement scores are associated with better test performance, thus indicating predictive validity of the engagement survey.</p> <p>Table 1 shows the engagement scores and test performance by conditions. We conducted paired sample <emph>t</emph>‐tests, using instructor visibility as the independent variable and engagement scores and proportion correct on the test as dependent variables. Participants reported slightly higher emotional engagement after watching the instructor‐visible lecture than the instructor‐not‐visible lecture, <emph>t</emph>(<reflink idref="bib143" id="ref113">143</reflink>) = 1.99, <emph>p</emph> = 0.048, <emph>d</emph> = 0.17, <emph>B</emph><subs>10</subs> = 0.63. Note that although this comparison was significant in NHST, the Bayes factor was inconclusive. For behavioral engagement, the difference was not significant, <emph>t</emph>(<reflink idref="bib143" id="ref114">143</reflink>) = 1.11, <emph>p</emph> = 0.271, <emph>d</emph> = 0.09, <emph>B</emph><subs>01</subs> = 5.93. Lastly, instructor visibility did not influence learning as measured by test performance, <emph>t</emph>(<reflink idref="bib143" id="ref115">143</reflink>) = 0.65, <emph>p</emph> = 0.515, <emph>d</emph> = 0.05, <emph>B</emph><subs>01</subs> = 8.75.</p> <p>1 TABLE Engagement score and proportion of test performance in Experiment 1.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Condition</th><th align="center">Engagement</th><th align="center">Test performance</th></tr><tr><th align="center">Emotional</th><th align="center">Behavioral</th></tr><tr><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th></tr></thead><tbody valign="top"><tr><td align="left">Instructor‐visible</td><td align="center">3.10</td><td align="center">0.93</td><td align="center">3.64</td><td align="center">0.82</td><td align="center">0.59</td><td align="center">0.25</td></tr><tr><td align="left">Instructor‐not‐visible</td><td align="center">2.93</td><td align="center">0.84</td><td align="center">3.57</td><td align="center">0.83</td><td align="center">0.57</td><td align="center">0.24</td></tr></tbody></table> </ephtml> </p> <hd id="AN0184623579-20">Effects of Individual Preference</hd> <p>Overall, the majority of participants preferred instructor‐visible videos (50% in general preference and 51% in experimental preference), but a considerable minority preferred instructor‐not‐visible videos (24% in general and 32% in experimental), with the remaining indicating no preference (26% in general and 17% in experimental).</p> <p>To determine the extent to which individual preference for instructor visibility affected engagement, we conducted a 3 (general preference: visible‐preferred, not‐visible‐preferred, no preference) × 2 (instructor visibility: visible, not‐visible) repeated measures ANOVA, with engagement score serving as the dependent variable. For simplicity, we used the total engagement score by averaging the emotional and behavioral engagement scores. The same ANOVA was conducted with experimental preference as well.</p> <p>Of the two analyses, only the interaction between <emph>experimental preference</emph> and instructor visibility was significant, <emph>F</emph>(<reflink idref="bib2" id="ref116">2</reflink>, 141) = 13.92, <emph>p</emph> < 0.001, <emph>η</emph><subs>p</subs><sups>2</sups> = 0.05, <emph>B</emph><subs>10</subs> = 7.87. Table 2 shows the relevant data. When participants preferred the instructor‐visible video, they also provided higher engagement ratings for the instructor‐visible lecture than the instructor‐not‐visible lecture, <emph>t</emph>(<reflink idref="bib73" id="ref117">73</reflink>) = 4.54, <emph>p</emph> < 0.001, <emph>d</emph> = 0.53, <emph>B</emph><subs>10</subs> = 845.76. In contrast, when participants preferred the instructor‐not‐visible video, they reported greater engagement with the instructor‐not‐visible lecture than in the instructor‐visible lecture, <emph>t</emph>(<reflink idref="bib45" id="ref118">45</reflink>) = −3.96, <emph>p</emph> < 0.001, <emph>d</emph> = −0.58, <emph>B</emph><subs>10</subs> = 96.85. Lastly, when participants reported no preference, their engagement ratings did not differ significantly between the lectures, <emph>t</emph>(<reflink idref="bib23" id="ref119">23</reflink>) = 0.88, <emph>p</emph> = 0.390, <emph>d</emph> = 0.18, <emph>B</emph><subs>01</subs> = 3.30. These data showed that participants were internally consistent when they reported their preference and engagement toward the lectures. We also conducted the same analyses using test performance as the dependent variable to examine whether individual preference affected learning (see Table 2 for means). It did not, <emph>p</emph>s > 0.217.</p> <p>2 TABLE Engagement score and test performance as a function of instructor visibility and experimental preference in Experiment 1.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Condition</th><th align="center">Engagement</th><th align="center">Test performance</th></tr><tr><th align="center">Visible‐preferred (<italic>N</italic> = 74)</th><th align="center">Not‐visible‐preferred (<italic>N</italic> = 46)</th><th align="center">No‐preference (<italic>N</italic> = 24)</th><th align="center">Visible‐preferred (<italic>N</italic> = 74)</th><th align="center">Not‐visible‐preferred (<italic>N</italic> = 46)</th><th align="center">No‐preference (<italic>N</italic> = 24)</th></tr><tr><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th></tr></thead><tbody valign="top"><tr><td align="left">Instructor‐visible</td><td align="center">3.54</td><td align="center">0.76</td><td align="center">3.19</td><td align="center">0.67</td><td align="center">3.23</td><td align="center">0.96</td><td align="center">0.62</td><td align="center">0.27</td><td align="center">0.60</td><td align="center">0.26</td><td align="center">0.57</td><td align="center">0.24</td></tr><tr><td align="left">Instructor‐not‐visible</td><td align="center">3.15</td><td align="center">0.69</td><td align="center">3.53</td><td align="center">0.69</td><td align="center">3.05</td><td align="center">0.82</td><td align="center">0.57</td><td align="center">0.20</td><td align="center">0.63</td><td align="center">0.28</td><td align="center">0.50</td><td align="center">0.26</td></tr></tbody></table> </ephtml> </p> <hd id="AN0184623579-21">Reasons Underlying Instructor Visibility Preference</hd> <p>To analyze participants' reasons for their <emph>experimental</emph> preference for instructor visibility, we coded their written responses using OpenAI's GPT 4 (OpenAI [<reflink idref="bib57" id="ref120">57</reflink>]).[<reflink idref="bib3" id="ref121">3</reflink>] Because participants who reported <emph>no preference</emph> were not asked to justify their decisions, only those who indicated a clear preference for either instructor‐visible or instructor‐not‐visible were considered for this analysis.</p> <p>First, we explained the context of the current study to GPT and asked it to categorize the reasons that led participants to prefer the instructor‐visible video. Specifically, we provided the following prompt: <emph>I'll provide the students' responses about why they prefer to see the instructor's image in the online lecture videos. Could you categorize them into multiple groups and explain each category?</emph> GPT identified eight categories based on this prompt, which are available on our OSF page. Using this initial categorization as a foundation, the first author refined the resulting categories by merging similar ones, removing redundant ones, and creating a new one. We followed the same procedure to categorize the written responses for participants who preferred the instructor‐not‐visible video.</p> <p>For participants who preferred the instructor‐visible video (<emph>N</emph> = 74), the responses were categorized as follows:</p> <p></p> <ulist> <item> Personal connection/realism—a sense of connectedness, affective response, or engagement with the lecture/instructor, or realness of a lecture.</item> <p></p> <item> Attention—easy to focus and pay attention to.</item> <p></p> <item> Visual cues—able to see the instructor's facial expression, lip movements, or gestures.</item> <p></p> <item> Comprehension/retention—able to understand or retain the material better.</item> <p></p> <item> Topic interest—the topic was more interesting or relevant.</item> </ulist> <p>For participants who preferred the instructor‐not‐visible video (<emph>N</emph> = 46), GPT identified seven categories of reasoning, which we refined and condensed into three categories as follows:</p> <p></p> <ulist> <item> Distraction—the presence of instructor's face can be distracting or easy to focus without it.</item> <p></p> <item> Topic interest—the topic was more interesting or relevant, or participants did not like the instructor‐visible lecture video.</item> <p></p> <item> Miscellaneous—any other reasons not mentioned above.</item> </ulist> <p>Based on this categorization, we instructed GPT to classify each participant's response into the relevant categories. We specified that participants' responses could include multiple categories. Examples of responses for each category are presented in the Supporting Information. Details of the GPT prompt used, along with all participant responses and their categorizations, are also available on the OSF page.</p> <p>Figure 3 shows the distribution of reasons for preferring instructor‐visible videos. Most participants mentioned personal connection/realism and attention as their justification for preferring the instructor‐visible video, with visual aids and comprehension as other noteworthy categories. The justifications for preferring instructor‐not‐visible videos are shown in Figure 4, in which the vast majority of participants cited distraction as their primary reason. For both preferences—whether instructor‐visible or ‐not‐visible—some participants noted that their interest in the subject matter influenced their choice, which was independent of instructor visibility.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/BU8/01mar25/acp70026-fig-0003.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="acp70026-fig-0003.jpg" title="3 Distribution of preference reasons for instructor‐visible videos. As a participant's response could fall into multiple categories, the sum of percentage can be greater than 100%." /> </p> <p></p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/BU8/01mar25/acp70026-fig-0004.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="acp70026-fig-0004.jpg" title="4 Distribution of preference reasons for instructor‐not‐visible videos. As a participant's response could fall into multiple categories, the sum of percentage can be greater than 100%." /> </p> <p></p> <hd id="AN0184623579-24">Discussion</hd> <p>As expected, test performance did not differ between the instructor‐visible and instructor‐not‐visible conditions. Participants reported greater emotional engagement when they watched a lecture video with the instructor visible, but behavioral engagement did not differ regardless of instructor visibility. Given that the positive effect of instructor visibility on emotional engagement was small and the <emph>p</emph> value was close to 0.05, we aimed to replicate this finding in Experiment 2 with a larger sample.</p> <p>Perhaps the most important finding from Experiment 1 was that participants found the lecture more engaging when it aligned with their instructor preference. Those who favored instructor‐visible videos exhibited greater engagement with instructor‐visible videos compared to instructor‐not‐visible ones, and the opposite was true for participants who favored instructor‐not‐visible videos. However, this effect was shown only for experimental preference rather than general preference, which constrains the finding's generalizability. Alternatively, it is possible that our general preference measure did not accurately capture participants' everyday preferences. Because we measured general preference <emph>after</emph> participants had viewed the two lecture videos, their responses might have reflected their recent video‐watching experiences rather than their true preferences. For example, a participant who normally prefers to see the instructor might dislike the experiment's lecture video with the instructor visible—perhaps the participant did not enjoy the topic or the specific delivery by the instructor—and let this experience alter their answer to the general preference question. Indeed, a considerable number of participants (<emph>N</emph> = 20) reported topic interest as their reason for preferring either instructor‐visible or ‐not‐visible videos. To eliminate this potential issue, participants completed the preference survey at the beginning of Experiment 2.</p> <hd id="AN0184623579-25">Experiment 2</hd> <p>In Experiment 2, we also examined whether video playback speed, another important attribute in online lectures, affected engagement and learning. Students often report speeding up video playback while watching lectures (e.g., 1.5× speed, Lang et al. [<reflink idref="bib43" id="ref122">43</reflink>]; Murphy et al. [<reflink idref="bib53" id="ref123">53</reflink>]). Research has also shown that people often find moderately fast speakers more engaging compared to slow speakers (Guo, Kim, and Rubin [<reflink idref="bib30" id="ref124">30</reflink>]; Perloff [<reflink idref="bib61" id="ref125">61</reflink>]; Simonds et al. [<reflink idref="bib74" id="ref126">74</reflink>]). Therefore, increasing video playback speed may have a positive effect on engagement. Indeed, Murphy, Hoover, and Castel ([<reflink idref="bib54" id="ref127">54</reflink>]) reported a reduction in mind‐wandering when individuals watched online lectures at faster playback speeds. Prior studies have also explored the impact of video playback speed on learning, although the findings are somewhat mixed. For instance, Nagahama and Morita ([<reflink idref="bib55" id="ref128">55</reflink>]) had undergraduate students watch a video lecture at different playback speeds (1.0×, 1.5×, or 2.0×) and found that test performance was best in the 1.5× condition. In contrast, Song et al. ([<reflink idref="bib79" id="ref129">79</reflink>]) reported that participants' test performance was impaired when they watched a lecture at 1.5× speed compared with 1.0× speed.</p> <p>We were particularly interested in examining the potential interaction between instructor visibility and video playback speed for learning. Seeing the speaker's lip movements or gestures can enhance speech comprehension (Schwartz, Berthommier, and Savariaux [<reflink idref="bib70" id="ref130">70</reflink>]; Skipper, Nusbaum, and Small [<reflink idref="bib76" id="ref131">76</reflink>]; Sueyoshi and Hardison [<reflink idref="bib80" id="ref132">80</reflink>]), especially when listeners face challenging situations (Piquard‐Kipffer et al. [<reflink idref="bib63" id="ref133">63</reflink>]; Shindo, Kaga, and Tanaka [<reflink idref="bib73" id="ref134">73</reflink>]). For instance, Dahl and Ludvigsen ([<reflink idref="bib18" id="ref135">18</reflink>]) reported that speaker gestures in lecture videos improved recall and comprehension, but only for those who speak English as a foreign language. Similarly, Piquard‐Kipffer et al. ([<reflink idref="bib63" id="ref136">63</reflink>]) reported that children at risk of reading failure performed better in listening tasks when they could observe a speaker's lip movements. Although watching a video at a fast speed is not the same as watching a speaker as a second language learner or as an individual with reading difficulty, both scenarios share a commonality: individuals may find it challenging to follow the video material. In this context, instructor visibility might have a more positive effect at faster playback speeds (1.5× and 2.0×) compared with regular playback speed (1.0×).</p> <hd id="AN0184623579-26">Design and Participants</hd> <p>We manipulated instructor visibility within‐subjects (instructor‐visible vs. instructor‐not‐visible) and playback speed between‐subjects (1.0× vs. 1.5× vs. 2.0×). We aimed to triple the goal sample size of Experiment 1 (<emph>N</emph> = 420, 140 × 3), as there were three levels in the playback speed variable. Participants were 445 undergraduate students from a large Midwest university who completed the study for course credits. Data from 22 participants were excluded from analyses for pre‐registered criteria, which are detailed in the Supporting Information. The final sample included data from 423 participants (253 women, 165 men, two non‐binary, two did not respond, and one transgender; mean age = 19.23), with 140 in the 1.0× condition, 142 in the 1.5× condition, and 141 in the 2.0× condition.</p> <hd id="AN0184623579-27">Materials and Procedure</hd> <p>The materials and procedure were the same as in Experiment 1 except for the following changes. First, we manipulated playback speed (1.0×, 1.5×, and 2.0×). Second, given our interest in general preference, we removed the question regarding experimental preference, and the preference survey was conducted at the beginning. We again included a categorical question (general preference) and quantitative questions (instructor preference rating). Third, questions regarding video playback speed preference were added to the preference survey. Specifically, participants were asked the following question, "When watching online video lectures, which playing speed do you prefer?" and offered five options: <emph>slower than 1.0×, 1.0×, 1.5×, 2.0×, and faster than 2.0×</emph>. This was followed by a justification question, "Could you tell us why you prefer the speed you chose above?" Additionally, three Likert‐scale items regarding speed preference were added to the survey. The quantitative measure of speed preference is detailed in the Supporting Information. Like Experiment 1, 26 participants reported that they were either extremely or very familiar with the material, but including these participants did not change our conclusions, so we included their data in the analyses.</p> <hd id="AN0184623579-28">Results</hd> <p>Similar to Experiment 1, we found no interaction effects across lecture topics, instructor gender, instructor visibility condition, and speed, for both engagement and test performance, <emph>F</emph>s < 1.78, <emph>p</emph>s > 0.099, <emph>η</emph><subs>p</subs><sups>2</sups>s < 0.01, <emph>B</emph><subs>01</subs>s > 1.04. Thus, the following analyses were conducted after collapsing the data across lecture topics and instructor genders.</p> <hd id="AN0184623579-29">Effects of Instructor Visibility and Playback Speed on Engagement and Learning</hd> <p>The engagement survey demonstrated good reliability, consistent with Experiment 1 (Cronbach's <emph>α</emph> = 0.85 for emotional engagement and 0.81 for behavioral engagement). There was also a positive correlation between engagement score and test performance in both the instructor‐visible condition, (<emph>r</emph><subs>emotional</subs>[<reflink idref="bib421" id="ref137">421</reflink>] = 0.36, <emph>p</emph> < 0.001, <emph>B</emph><subs>10</subs> = 1.565 × 10<sups>11</sups>; <emph>r</emph><subs>behavioral</subs>[<reflink idref="bib421" id="ref138">421</reflink>] = 0.33, <emph>p</emph> < 0.001, <emph>B</emph><subs>10</subs> = 1.864 × 10<sups>9</sups>), and the instructor‐not‐visible condition, (<emph>r</emph><subs>emotional</subs>[<reflink idref="bib421" id="ref139">421</reflink>] = 0.35, <emph>p</emph> < 0.001, <emph>B</emph><subs>10</subs> = 5.956 × 10<sups>10</sups>; <emph>r</emph><subs>behavioral</subs>[<reflink idref="bib421" id="ref140">421</reflink>] = 0.33, <emph>p</emph> < 0.001, <emph>B</emph><subs>10</subs> = 2.992 × 10<sups>9</sups>).</p> <p>To examine whether instructor visibility and playback speed affected <emph>engagement</emph>, we conducted a 2 (instructor visibility: visible, not‐visible) × 3 (speed: 1.0×, 1.5×, 2.0×) repeated‐measures ANOVA (see Table 3 for means) with emotional and behavioral engagement as the dependent variable, respectively. Emotional engagement score did not differ significantly between the instructor‐visible and instructor‐not‐visible conditions (<emph>M</emph><subs>visible</subs> = 3.03, <emph>M</emph><subs>not‐visible</subs> = 3.00), <emph>F</emph>(<reflink idref="bib1" id="ref141">1</reflink>, 420) = 0.22, <emph>p</emph> = 0.639, <emph>d</emph> = 0.02, <emph>B</emph><subs>01</subs> = 16.40, and neither did behavioral engagement (<emph>M</emph><subs>visible</subs> = 3.56, <emph>M</emph><subs>not‐visible</subs> = 3.50), <emph>F</emph>(<reflink idref="bib1" id="ref142">1</reflink>, 420) = 2.37, <emph>p</emph> = 0.124, <emph>d</emph> = 0.08, <emph>B</emph><subs>01</subs> = 5.61. There was also no main effect of speed on emotional engagement (<emph>M</emph><subs>1.0×</subs> = 2.92, <emph>M</emph><subs>1.5×</subs> = 3.12, <emph>M</emph><subs>2.0×</subs> = 3.00), <emph>F</emph>(<reflink idref="bib2" id="ref143">2</reflink>, 420) = 2.90, <emph>p</emph> = 0.056, <emph>η</emph><subs>p</subs><sups>2</sups> = 0.01, <emph>B</emph><subs>01</subs> = 2.57, or behavioral engagement (<emph>M</emph><subs>1.0×</subs> = 3.51, <emph>M</emph><subs>1.5×</subs> = 3.59, <emph>M</emph><subs>2.0×</subs> = 3.50), <emph>F</emph>(<reflink idref="bib2" id="ref144">2</reflink>, 420) = 0.75, <emph>p</emph> = 0.472, <emph>η</emph><subs>p</subs><sups>2</sups> = 0.00, <emph>B</emph><subs>01</subs> = 18.90. Interaction effects were not significant, either, <emph>F</emph>s < 0.11, <emph>p</emph>s > 0.899, <emph>η</emph><subs>p</subs><sups>2</sups>s < 0.01, <emph>B</emph><subs>01</subs>s > 34.05.</p> <p>3 TABLE Engagement score as a function of instructor visibility and playback speed in Experiment 2.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Condition</th><th align="center">Emotional engagement</th><th align="center">Behavioral engagement</th></tr><tr><th align="center">1.0×</th><th align="center">1.5×</th><th align="center">2.0×</th><th align="center">1.0×</th><th align="center">1.5×</th><th align="center">2.0×</th></tr><tr><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th></tr></thead><tbody valign="top"><tr><td align="left">Instructor‐visible</td><td align="center">2.93</td><td align="center">0.92</td><td align="center">3.14</td><td align="center">0.87</td><td align="center">3.01</td><td align="center">0.89</td><td align="center">3.53</td><td align="center">0.74</td><td align="center">3.63</td><td align="center">0.77</td><td align="center">3.52</td><td align="center">0.83</td></tr><tr><td align="left">Instructor‐not‐visible</td><td align="center">2.90</td><td align="center">0.89</td><td align="center">3.10</td><td align="center">0.90</td><td align="center">3.00</td><td align="center">0.84</td><td align="center">3.49</td><td align="center">0.79</td><td align="center">3.55</td><td align="center">0.81</td><td align="center">3.47</td><td align="center">0.86</td></tr></tbody></table> </ephtml> </p> <p>We now report the results of the same 2 × 3 repeated measures ANOVA but with <emph>test performance</emph> as the dependent variable (see Table 4 for means). Consistent with the results in Experiment 1, test performance was not influenced by instructor visibility (<emph>M</emph><subs>visible</subs> = 0.56, <emph>M</emph><subs>not‐visible</subs> = 0.57), <emph>F</emph>(<reflink idref="bib1" id="ref145">1</reflink>, 420) = 0.08, <emph>p</emph> = 0.783, <emph>d</emph> = −0.01, <emph>B</emph><subs>01</subs> = 12.49. However, there was a main effect of playback speed (<emph>M</emph><subs>1.0×</subs> = 0.57, <emph>M</emph><subs>1.5×</subs> = 0.59, <emph>M</emph><subs>2.0×</subs> = 0.53), <emph>F</emph>(<reflink idref="bib2" id="ref146">2</reflink>, 420) = 3.59, <emph>p</emph> = 0.029, <emph>η</emph><subs>p</subs><sups>2</sups> = 0.01, <emph>B</emph><subs>10</subs> = 0.41. Pairwise comparisons revealed that participants in the 1.5× condition showed superior test performance relative to those in the 2.0× condition, <emph>t</emph>(<reflink idref="bib281" id="ref147">281</reflink>) = 2.70, <emph>p</emph> = 0.007, <emph>d</emph> = 0.32, <emph>B</emph><subs>10</subs> = 4.04, with the remaining pairwise comparisons not showing significance, <emph>t</emph>s < 1.44, <emph>p</emph>s > 0.155, <emph>d</emph>s < 0.18, <emph>B</emph><subs>01</subs>s > 2.90. The interaction between instructor visibility and playback speed was not significant, <emph>F</emph>(<reflink idref="bib2" id="ref148">2</reflink>, 420) = 1.13, <emph>p</emph> = 0.324, <emph>η</emph><subs>p</subs><sups>2</sups> = 0.00, <emph>B</emph><subs>01</subs> = 12.22.</p> <p>4 TABLE Test performance as a function of instructor visibility and playback speed in Experiment 2.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Condition</th><th align="center">Test performance</th></tr><tr><th align="center">1.0×</th><th align="center">1.5×</th><th align="center">2.0×</th></tr><tr><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th></tr></thead><tbody valign="top"><tr><td align="left">Instructor‐visible</td><td align="center">0.56</td><td align="center">0.27</td><td align="center">0.57</td><td align="center">0.25</td><td align="center">0.55</td><td align="center">0.25</td></tr><tr><td align="left">Instructor‐not‐visible</td><td align="center">0.57</td><td align="center">0.26</td><td align="center">0.61</td><td align="center">0.25</td><td align="center">0.52</td><td align="center">0.24</td></tr></tbody></table> </ephtml> </p> <hd id="AN0184623579-30">Effects of Individual Preference</hd> <p>Amongst our participants, 54% reported a general preference for instructor‐visible videos, 15% for instructor‐not‐visible videos, and 31% reported no preference. For playback speed, very few participants noted that they preferred slower than 1.0× (1%) and faster than 2.0× (1%). Thus, we merged these options with their nearest one. The majority of participants favored 1.0× (44%) and 1.5× (49%), and a small minority favored 2.0× (8%).</p> <p>Similar to Experiment 1's experimental preference results, we observed an interaction between individual preference and instructor visibility on total engagement score—except that these results are about participants' <emph>general</emph> preference, <emph>F</emph>(<reflink idref="bib2" id="ref149">2</reflink>, 420) = 3.28, <emph>p</emph> = 0.039, <emph>η</emph><subs>p</subs><sups>2</sups> = 0.004, <emph>B</emph><subs>10</subs> = 0.57 (see Table 5). Specifically, participants who preferred to see the instructor found the instructor‐visible video more engaging, <emph>t</emph>(<reflink idref="bib230" id="ref150">230</reflink>) = 1.72, <emph>p</emph> = 0.087, <emph>d</emph> = 0.11, <emph>B</emph><subs>10</subs> = 0.31, although this difference was not significant, and the opposite occurred for participants who preferred not to see the instructor, <emph>t</emph>(<reflink idref="bib61" id="ref151">61</reflink>) = −2.03, <emph>p</emph> = 0.046, <emph>d</emph> = −0.26, <emph>B</emph><subs>10</subs> = 0.95. Finally, those who reported no preference found the two lecture videos similarly engaging, <emph>t</emph>(<reflink idref="bib129" id="ref152">129</reflink>) = 0.94, <emph>p</emph> = 0.347, <emph>d</emph> = 0.08, <emph>B</emph><subs>01</subs> = 6.64.</p> <p>5 TABLE Engagement score and test performance as a function of instructor visibility and general preference in Experiment 2.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Condition</th><th align="center">Engagement</th><th align="center">Test performance</th></tr><tr><th align="center">Visible‐preferred (<italic>N</italic> = 231)</th><th align="center">Not‐visible‐preferred (<italic>N</italic> = 62)</th><th align="center">No‐preference (<italic>N</italic> = 130)</th><th align="center">Visible‐preferred (<italic>N</italic> = 231)</th><th align="center">Not‐visible‐preferred (<italic>N</italic> = 62)</th><th align="center">No‐preference (<italic>N</italic> = 130)</th></tr><tr><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th></tr></thead><tbody valign="top"><tr><td align="left">Instructor‐visible</td><td align="center">3.37</td><td align="center">0.78</td><td align="center">3.20</td><td align="center">0.72</td><td align="center">3.20</td><td align="center">0.71</td><td align="center">0.56</td><td align="center">0.25</td><td align="center">0.54</td><td align="center">0.25</td><td align="center">0.57</td><td align="center">0.27</td></tr><tr><td align="left">Instructor‐not‐visible</td><td align="center">3.28</td><td align="center">0.73</td><td align="center">3.40</td><td align="center">0.78</td><td align="center">3.13</td><td align="center">0.81</td><td align="center">0.58</td><td align="center">0.25</td><td align="center">0.57</td><td align="center">0.25</td><td align="center">0.54</td><td align="center">0.26</td></tr></tbody></table> </ephtml> </p> <p>We also examined if participants' preference for playback speed interacted with the effect of playback speed on engagement and learning. It did not, all <emph>p</emph>s > 0.128. The relevant means are shown in Table 6.</p> <p>6 TABLE Engagement score and test performance as a function of playback speed and preference in Experiment 2.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Condition</th><th align="center">Engagement</th><th align="center">Test performance</th></tr><tr><th align="center">1.0× preferred (<italic>N</italic> = 184)</th><th align="center">1.5× preferred (<italic>N</italic> = 205)</th><th align="center">2.0× preferred (<italic>N</italic> = 34)</th><th align="center">1.0× preferred (<italic>N</italic> = 184)</th><th align="center">1.5× preferred (<italic>N</italic> = 205)</th><th align="center">2.0× preferred (<italic>N</italic> = 34)</th></tr><tr><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th><th align="center"><italic>M</italic></th><th align="center">SD</th></tr></thead><tbody valign="top"><tr><td align="left">1.0×</td><td align="center">3.26</td><td align="center">0.65</td><td align="center">3.21</td><td align="center">0.63</td><td align="center">2.98</td><td align="center">0.66</td><td align="center">0.55</td><td align="center">0.17</td><td align="center">0.57</td><td align="center">0.20</td><td align="center">0.60</td><td align="center">0.16</td></tr><tr><td align="left">1.5×</td><td align="center">3.46</td><td align="center">0.68</td><td align="center">3.27</td><td align="center">0.58</td><td align="center">3.42</td><td align="center">0.63</td><td align="center">0.58</td><td align="center">0.18</td><td align="center">0.60</td><td align="center">0.18</td><td align="center">0.61</td><td align="center">0.19</td></tr><tr><td align="left">2.0×</td><td align="center">3.17</td><td align="center">0.74</td><td align="center">3.31</td><td align="center">0.56</td><td align="center">3.41</td><td align="center">0.50</td><td align="center">0.52</td><td align="center">0.19</td><td align="center">0.56</td><td align="center">0.18</td><td align="center">0.47</td><td align="center">0.20</td></tr></tbody></table> </ephtml> </p> <hd id="AN0184623579-31">Reasons Underlying Instructor Visibility Preference</hd> <p>To categorize the reasons that underlay participants' preference for seeing the instructor or not, we again used GPT to classify participants' written responses based on the categories established during Experiment 1. However, because we asked participants in Experiment 2 to justify their <emph>general</emph> preference instead of <emph>experimental</emph> preference, topic interest was omitted from the categorization. Aside from this exception, we employed the same methodology as Experiment 1 to categorize participants' responses.</p> <p>The results largely replicated those from Experiment 1 (see Figure 3). Many participants cited personal connection/realism and attention as their reasons for preferring to see the instructor, while a considerable proportion mentioned comprehension and visual aids. In contrast, for participants who preferred instructor‐not‐visible videos, nearly all cited distraction as their reason, as shown in Figure 4.</p> <hd id="AN0184623579-32">Reasons Underlying Playback Speed Preference</hd> <p>We categorized participants' justifications for playback speed in the same manner as their instructor preference. Specifically, we first directed GPT to generate categories based on the text input of participants who preferred 1.0× speed. From this initial categorization, the first author further refined the categories for the 1.0× speed preference as follows:</p> <p></p> <ulist> <item> Natural pace—it is the normal, regular speed, or it feels authentic.</item> <p></p> <item> Comprehension/retention—it is ideal for understanding or retaining the material presented.</item> <p></p> <item> Note‐taking—it gives enough time to take notes without feeling rushed. Participants also mentioned concerns regarding missing important information.</item> <p></p> <item> Attention—it is easier to focus and pay attention.</item> <p></p> <item> Miscellaneous—any other reasons not mentioned above.</item> </ulist> <p>For the 1.5× speed (<emph>N</emph> = 205), we repeated the same procedure as with the 1.0× speed categorization, and the reasons were categorized as follows:</p> <p></p> <ulist> <item> Efficiency/time management—participants can manage their time better or they can still understand the material at 1.5× speed.</item> <p></p> <item> Optimality—the optimality of speed, such as being fast but not too fast.</item> <p></p> <item> Attention—it is easier to focus and pay attention.</item> <p></p> <item> Control—participants can pause or stop videos at any time.</item> <p></p> <item> Miscellaneous—any other reasons not mentioned above.</item> </ulist> <p>Participants who preferred 2.0× speed (<emph>N</emph> = 34) reported similar reasons to those who preferred 1.5× speed, with the exception of the optimality category. Consequently, we used the same categories for 2.0× speed as for 1.5× speed, except for the optimality category.</p> <p>Figures 5 and 6 display the distribution of preference reasons for the regular 1.0× speed and the faster speeds (i.e., 1.5× and 2.0×), respectively. For participants who preferred 1.0× speed, natural pace was cited as the primary reason, with comprehension/retention and note‐taking being other common reasons. For those who preferred to watch video lectures at 1.5× and 2.0× speed, efficiency/time management and attention made up the bulk of their justifications.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/BU8/01mar25/acp70026-fig-0005.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="acp70026-fig-0005.jpg" title="5 Distribution of preference reasons 1.0× speed videos. As a participant's response could fall into multiple categories, the sum of percentages can be greater than 100%." /> </p> <p></p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/BU8/01mar25/acp70026-fig-0006.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="acp70026-fig-0006.jpg" title="6 Distribution of preference reasons 1.5× and 2.0× speed videos. As a participant's response could fall into multiple categories, the sum of percentages can be greater than 100%." /> </p> <p></p> <hd id="AN0184623579-35">Discussion</hd> <p>Instructor visibility benefited emotional engagement in Experiment 1, but it did not impact either emotional or behavioral engagement in Experiment 2. We do not believe that these null effects can be attributed to the additional playback speed manipulation in Experiment 2, as there was no interaction between video playback speed and instructor visibility. Furthermore, this null effect of instructor visibility on engagement was still found when the analyses were restricted to participants in the 1.0× condition, <emph>t</emph>s < 0.66, <emph>p</emph>s > 0.509, <emph>d</emph>s < 0.06, <emph>B</emph><subs>01</subs> > 8.58. Thus, the increase in emotional engagement for instructor‐visible over instructor‐not‐visible video in Experiment 1 is likely due to a sampling error, especially given that Experiment 2's sample size is tripled that of Experiment 1's.</p> <p>More importantly, we found that participants reported the lecture as more engaging when it aligned with their preference for instructor visibility. This effect was shown in Experiment 1 with <emph>experimental</emph> preference, and the same pattern was again observed with <emph>general</emph> preference in Experiment 2. The consistency of this finding across two experiments is particularly noteworthy because (i) one was about a prevailing preference and the other was about a transient preference, and (ii) we moved the preference survey to the beginning of Experiment 2 so that we could rule out the possibility that the video‐watching experience influenced participants' responses. Thus, the preference measure in Experiment 2 likely captured participants' everyday preference for online lectures.</p> <p>Consistent with the results of Experiment 1, instructor visibility did not affect learning. Indeed, participants produced nearly identical performance regardless of whether the instructor appeared in the lecture (<emph>M</emph><subs>visible</subs> = 0.56, <emph>M</emph><subs>not‐visible</subs> = 0.57). In contrast, playback speed affected learning but not engagement. Specifically, watching the lecture at 1.5× speed yielded the highest test performance (<emph>M</emph> = 0.59), whereas watching the video at 2.0× speed resulted in the lowest (<emph>M</emph> = 0.53). Although there was a slight reduction in test score for participants who watched the lecture at 2.0× compared with 1.0× (<emph>M</emph> = 0.57), this drop was modest especially when considering the time saved. Indeed, when we consider the test performance and time saved in the 2.0× and especially the 1.5× conditions—watching a lecture video at a faster speed could be considered a <emph>more efficient</emph> method of learning.</p> <p>Lastly, we were also interested in whether there was an interaction between instructor visibility and video playback speed for learning outcomes. However, no such interaction was observed. To explore whether a more extreme speed difference would reveal an interaction, we conducted an analysis including only the 1.0× and 2.0× speed conditions. The interaction remained non‐significant, with <emph>F</emph>(<reflink idref="bib1" id="ref153">1</reflink>, 279) = 0.61, <emph>p</emph> = 0.434, <emph>η</emph><subs>p</subs><sups>2</sups> = 0.00, <emph>B</emph><subs>01</subs> = 5.81, suggesting that the impact of instructor visibility on video playback speed was negligible.</p> <hd id="AN0184623579-36">General Discussion</hd> <p>In two experiments, we investigated the effects of instructor visibility and playback speed on student engagement and learning, and whether individual preference interacted with these relationships. A strength of the current study is its rigorous experimental design (cf., Guo, Kim, and Rubin [<reflink idref="bib30" id="ref154">30</reflink>]; Lackmann et al. [<reflink idref="bib41" id="ref155">41</reflink>]; Pi, Hong, and Yang [<reflink idref="bib62" id="ref156">62</reflink>]). Furthermore, our pre‐registered experiments had sample sizes large enough to detect very small effect sizes (cf., Kizilcec, Papadopoulos, and Sritanyaratana [<reflink idref="bib36" id="ref157">36</reflink>]; van Wermeskerken and van Gog [<reflink idref="bib88" id="ref158">88</reflink>]; Wang and Antonenko [<reflink idref="bib90" id="ref159">90</reflink>]). Lastly, we used multiple lecture videos across a range of topics featuring various instructors, further enhancing the generalizability of our findings. Below, we summarize our key findings and discuss their implications.</p> <hd id="AN0184623579-37">Instructor Visibility and Engagement</hd> <p>Instructor visibility did not universally enhance participants' overall engagement. Although instructor visibility had a positive effect on emotional engagement in Experiment 1, this effect was small and was not replicated in Experiment 2. Our results appear to contradict previous findings that reported a positive impact of instructor visibility on perceived learning and affect (Kizilcec, Papadopoulos, and Sritanyaratana [<reflink idref="bib36" id="ref160">36</reflink>]; Kizilcec, Bailenson, and Gomez [<reflink idref="bib35" id="ref161">35</reflink>]; Wang and Antonenko [<reflink idref="bib90" id="ref162">90</reflink>]; Wilson et al. [<reflink idref="bib94" id="ref163">94</reflink>]). However, an alternative interpretation is that instructor visibility increased engagement in some studies because a majority of students in these studies preferred videos with an instructor visible. In our study, slightly over half of the participants preferred lecture videos that included the instructor's face. Similarly, Kizilcec, Bailenson, and Gomez ([<reflink idref="bib35" id="ref164">35</reflink>]) found that 57% of students who enrolled in a MOOC chose to watch videos featuring the instructor's face. In other words, the positive effect of instructor visibility on engagement might be driven by a prevailing, individual preference for instructor‐visible videos, rather than instructor visibility having an overarching impact on engagement.</p> <p>We consistently found that the impact of instructor visibility on engagement largely depended on individual preference. Specifically, participants found the video lecture more engaging when it matched their preference for instructor visibility, whereas those without a clear preference were unaffected. This effect was shown both with participants' experimental preference (Experiment 1) and general preference (Experiment 2), and when we asked for their preference before or after they watched the video lectures.</p> <p>Most prior research on instructor visibility has focused on its overall effect (Guo, Kim, and Rubin [<reflink idref="bib30" id="ref165">30</reflink>]; Homer, Plass, and Blake [<reflink idref="bib32" id="ref166">32</reflink>]; van Gog, Verveer, and Verveer [<reflink idref="bib87" id="ref167">87</reflink>]) without considering individual differences (cf. Kizilcec, Bailenson, and Gomez [<reflink idref="bib35" id="ref168">35</reflink>]). However, our results underscore the importance of taking into consideration student preferences when producing online lecture videos, given that the inclusion of an instructor can be <emph>detrimental</emph> to engagement for some students. Across 567 participants in our two experiments, 17% reported that they preferred not to see the instructor when watching lecture videos because they found the presence of an instructor distracting. This figure is relatively small compared with the 35% reported by Kizilcec and colleagues in their very large sample study. But 17% remain a sizable minority, so it might be worthwhile for instructors to provide students with two versions of a lecture video—one with the instructor visible and another without the instructor—to maximize engagement. If resources are constrained and producing two versions is challenging, we suggest creating the version with the instructor appearing at a fixed location on the screen, and informing students who find the instructor's image distracting to cover it with a sticky note.</p> <hd id="AN0184623579-38">Instructor Visibility and Learning</hd> <p>Across both experiments, instructor visibility did not promote learning. In some ways, this finding is not surprising given that previous studies have reported mixed results (Kizilcec, Papadopoulos, and Sritanyaratana [<reflink idref="bib36" id="ref169">36</reflink>]; van Gog, Verveer, and Verveer [<reflink idref="bib87" id="ref170">87</reflink>]; van Wermeskerken and van Gog [<reflink idref="bib88" id="ref171">88</reflink>]; Wilson et al. [<reflink idref="bib94" id="ref172">94</reflink>]). Proponents of positive effects argued that an instructor's image might provide a more personal touch and offer attentional cues to learners, whereas others suggested that instructor visibility introduces extraneous cognitive load irrelevant to the learning material.</p> <p>Recent studies have suggested several factors that can moderate the impact of instructor visibility on learning. For example, a meta‐analysis by Beege et al. ([<reflink idref="bib7" id="ref173">7</reflink>]) reported that instructor visibility had a positive effect on social science and informatics, but a negative effect on historical science.[<reflink idref="bib4" id="ref174">4</reflink>] Additionally, Sepp et al. ([<reflink idref="bib71" id="ref175">71</reflink>]) indicated that the behaviors of instructors in lecture videos, rather than mere visibility, may determine the effect. For instance, an instructor using gestures to highlight parts of a slide can help learners focus on the right material, enhancing learning. In contrast, an instructor merely being present without attention‐directing behaviors may not contribute to learning (cf. van Wermeskerken and van Gog [<reflink idref="bib88" id="ref176">88</reflink>]). The key takeaway is that instructor visibility alone is unlikely to enhance learning; instead, researchers should focus on the processes involved in specific learning tasks and how certain instructor behaviors might augment those processes.</p> <hd id="AN0184623579-39">Video Playback Speed and Learning</hd> <p>In Experiment 2, we explored the influence of playback speed on engagement and learning. Unlike instructor visibility, playback speed had a significant influence on learning but not engagement. Specifically, participants performed the best when they watched the lecture videos at 1.5× speed, followed by 1.0× speed and then 2.0× speed. On a raw performance level, these data suggest that learners should avoid playing back videos at 2.0× speed. However, the performance did not differ significantly between the 1.0× and 2.0× speed conditions, nor between the 1.0× and 1.5× speed conditions. Thus, when we consider these data from the perspective of learning <emph>efficiency</emph>, 1.5× and 2.0× speeds become more favorable. In line with this finding, recent studies have also reported that watching videos at a faster speed does not harm learning up to 1.5× (Risko, Liu, and Bianchi [<reflink idref="bib67" id="ref177">67</reflink>]) and 2.0× speed (Ashburner et al. [<reflink idref="bib5" id="ref178">5</reflink>]), although exceeding these speeds can decrease learning performance (Tran, Bianchi, and Risko [<reflink idref="bib86" id="ref179">86</reflink>]).</p> <p>A novel contribution of the current paper is our exploration of why students often opt for faster playback speed (cf. Tharumalingam and Risko [<reflink idref="bib85" id="ref180">85</reflink>]). We found that students strategically use faster playback speeds to enhance learning efficiency and improve attention. In particular, the substantial time saving is notable given that students often cite a lack of time as the primary reason to avoid effortful but effective learning strategies (Rea et al. [<reflink idref="bib64" id="ref181">64</reflink>]). Similar to this notion, Murphy et al. ([<reflink idref="bib53" id="ref182">53</reflink>]) argued that watching lecture videos at 2.0× speed twice could even improve test performance compared to watching videos at 1.0× speed only once. However, given that many students opt to speed up lecture playback speed to save time, it might be unrealistic to expect that students would rewatch a lecture. Still, students might be encouraged to use their saved time to engage in other learning‐enhancing activities, such as retrieval practice (Ahn and Chan [<reflink idref="bib3" id="ref183">3</reflink>]; Chan, Meissner, and Davis [<reflink idref="bib13" id="ref184">13</reflink>]; McDermott [<reflink idref="bib50" id="ref185">50</reflink>]; Risko, Liu, and Bianchi [<reflink idref="bib67" id="ref186">67</reflink>]).</p> <hd id="AN0184623579-40">Limitations and Future Directions</hd> <p>One limitation of the current study was that our lecture videos were quite short (~8 min), and everything occurred in a single session. Although recent studies on lecture video length suggest that students prefer shorter videos (under 10 min, Manasrah, Masoud, and Jaradat [<reflink idref="bib45" id="ref187">45</reflink>]; Ozan and Ozarslan [<reflink idref="bib58" id="ref188">58</reflink>]; under 6 min, Guo, Kim, and Rubin [<reflink idref="bib30" id="ref189">30</reflink>]), the traditional lecture videos in higher education are often far longer and persist across many weeks (e.g., 30 min to 1 h, Bauer, Malchow, and Meinel [<reflink idref="bib6" id="ref190">6</reflink>]; Lagerstrom, Johanes, and Ponsukcharoen [<reflink idref="bib42" id="ref191">42</reflink>]). Therefore, future research should investigate the effect of instructor visibility on engagement and learning using longer lecture video materials and across multiple study sessions. The effect of instructor visibility might be magnified in longer videos, especially when students must watch them across many classes.</p> <p>Another limitation is that some may argue that the alignment of engagement scores with individual preference merely shows that participants responded to the engagement survey in a manner consistent with their original preference. That is, individual preference might not be <emph>predictive</emph> of how instructor visibility would affect engagement. This interpretation is reasonable given that people often desire to appear consistent (Cialdini, Trost, and Newsom [<reflink idref="bib14" id="ref192">14</reflink>]; Falk and Zimmermann [<reflink idref="bib22" id="ref193">22</reflink>]). For this hypothesis to hold true, however, when participants complete the eight‐item engagement survey for the second lecture, they would need to recall their responses to the engagement survey for the first lecture and then adjust their answers accordingly (see Figure 2 for the procedure). This whole process would be mentally demanding as it requires remembering all prior responses. Given the low stakes of the current study (i.e., obtaining a course credit), we believe participants would not be motivated to exert the considerable effort required to appear consistent. As a future direction, researchers can use an objective measure of engagement such as video‐watching time—participants can pause the lecture when they feel bored and no longer want to engage with the lecture—such a procedure might alleviate the concerns against participants wanting to appear consistent when answering questionnaires.</p> <hd id="AN0184623579-41">Conclusions</hd> <p>The current findings suggest that instructor visibility alone might not have a significant impact on student engagement with and learning from a video lecture. However, our results also highlighted a more nuanced relationship between instructor visibility and engagement—that is, learners' engagement with video lectures might depend on their preference for instructor visibility. Given these findings, when resources permit, it is advisable to offer a version with and a version without the instructor visible.</p> <hd id="AN0184623579-42">Author Contributions</hd> <p> <bold>Dahwi Ahn:</bold> investigation, writing – original draft, writing – review and editing, data curation, conceptualization, methodology, project administration. <bold>Jason C. K. Chan:</bold> writing – review and editing, supervision, resources, conceptualization, funding acquisition, investigation.</p> <hd id="AN0184623579-43">Acknowledgments</hd> <p>This work is supported by the National Science Foundation under Grant 2017333. During the preparation of this work, the authors used OpenAI's GPT4 to analyze participants' reasons for their preference for instructor visibility and playback speed. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.</p> <hd id="AN0184623579-44">Ethics Statement</hd> <p>The current study was declared as exemption by Iowa State University IRB (21‐037) and followed US Federal Policy for the Protection of Human Subjects. All participants gave their informed consent prior to their inclusion in the study.</p> <hd id="AN0184623579-45">Conflicts of Interest</hd> <p>The authors declare no conflicts of interest.</p> <hd id="AN0184623579-46">Data Availability Statement</hd> <p>Both experiments were pre‐registered on the Open Science Framework (OSF) at https://osf.io/jd6v3 (Experiment 1) and https://osf.io/rc3pv (Experiment 2) prior to data collection. The data and Supporting Information are available on the project page (https://osf.io/xebwp/?view_only=78995d51ae1a4acda5002c021f0cf599).</p> <hd id="AN0184623579-47">A Appendix</hd> <p></p> <hd id="AN0184623579-48">Engagement Items</hd> <p></p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Type</th><th align="center">Item</th></tr></thead><tbody valign="top"><tr><td align="left">Behavioral engagement</td><td align="center">If I didn't understand, I gave up right away<xref ref-type="fn" rid="tfn1" /></td></tr><tr><td align="center">I put effort into learning the topic covered in the lecture</td></tr><tr><td align="center">I did other things when I was supposed to be paying attention<xref ref-type="fn" rid="tfn1" /></td></tr><tr><td align="center">I stayed focused</td></tr><tr><td align="left">Emotional engagement</td><td align="center">I wanted to understand what is learned in the topic covered in this lecture</td></tr><tr><td align="center">I felt good when I was partaking in this lecture</td></tr><tr><td align="center">I enjoyed learning new things about the topic covered in this lecture</td></tr><tr><td align="center">I think that the topic covered in this lecture is boring<xref ref-type="fn" rid="tfn1" /></td></tr></tbody></table> </ephtml> </p> <p>1 Reverse‐coded items.</p> <hd id="AN0184623579-49">B Appendix</hd> <p></p> <hd id="AN0184623579-50">Test Example Questions</hd> <p>An answer for each question is underlined.</p> <p><Computer Science></p> <p>What kind of knowledge is statements of fact?</p> <p></p> <ulist> <item> ○ Declarative knowledge</item> <p></p> <item> ○ Imperative knowledge</item> <p></p> <item> ○ Explicit Knowledge</item> <p></p> <item> ○ Tacit Knowledge</item> </ulist> <p>What kind of knowledge shows how to solve a problem?</p> <p></p> <ulist> <item> ○ Implicit Knowledge</item> <p></p> <item> ○ Imperative knowledge</item> <p></p> <item> ○ Explicit Knowledge</item> <p></p> <item> ○ Declarative knowledge</item> </ulist> <p><Ecology></p> <p>Where do toucans sleep?</p> <p></p> <ulist> <item> ○ Base of trees</item> <p></p> <item> ○ The grounds</item> <p></p> <item> ○ Tree holes</item> <p></p> <item> ○ Upper branches of trees</item> </ulist> <p>What is the average life span of a toucan in captivity?</p> <p></p> <ulist> <item> ○ 10 Years</item> <p></p> <item> ○ 29 Years</item> <p></p> <item> ○ 15 Years</item> <p></p> <item> ○ 18 years</item> </ulist> <p><Physics></p> <p>What happens when the flash tube goes off?</p> <p></p> <ulist> <item> ○ Photons are released from the electrons</item> <p></p> <item> ○ Electrons give off energy</item> <p></p> <item> ○ Photons move in the crystal</item> <p></p> <item> ○ Energy is pumped into the crystal</item> </ulist> <p>What is the normal state of an electron?</p> <p></p> <ulist> <item> ○ Ground state</item> <p></p> <item> ○ Level state</item> <p></p> <item> ○ Low level state</item> <p></p> <item> ○ Excited state</item> </ulist> <p><Statistics></p> <p>Descriptive statistics summarize the data, with values like maximums, minimums, and ___________.</p> <p></p> <ulist> <item> ○ Averages</item> <p></p> <item> ○ Errors</item> <p></p> <item> ○ Probabilities</item> <p></p> <item> ○ Proportions</item> </ulist> <p>Why do researchers use a sample rather than a population?</p> <p></p> <ulist> <item> ○ Population is not representative of individuals</item> <p></p> <item> ○ It is often not possible to collect a true population</item> <p></p> <item> ○ The sample is more accurate than the population</item> <p></p> <item> ○ The sample is more descriptive</item> </ulist> <p>GRAPH: Data S1 Supporting Information.</p> <p>GRAPH: Data S2 Supporting Information.</p> <p>GRAPH: Data S3 Supporting Information.</p> <ref id="AN0184623579-51"> <title> Footnotes </title> <blist> <bibl id="bib1" idref="ref16" type="bt">1</bibl> <bibtext> See https://data.bangtech.com/algorithm/switch%5ffrogs%5fto%5fthe%5fopposite%5fside.htm for a detailed description.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref10" type="bt">2</bibl> <bibtext> Learning outcome was not measured in this study.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref121" type="bt">3</bibl> <bibtext> While using GPT, we turned off the "Chat history & training" setting so the data was not shared with the server.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref11" type="bt">4</bibl> <bibtext> However, this negative effect was driven by only two studies, suggesting that caution might be necessary to interpret this finding.</bibtext> </blist> <blist> <bibl id="bib5" idref="ref6" type="bt">5</bibl> <bibtext> Funding: This work was supported by National Science Foundation, 2017333.</bibtext> </blist> </ref> <ref id="AN0184623579-52"> <title> References </title> <blist> <bibtext> Aguiar, E., N. V. Chawla, J. Brockman, G. A. Ambrose, and V. Goodrich. 2014. " Engagement vs Performance: Using Electronic Portfolios to Predict First Semester Engineering Student Retention." In Proceedings of the Fourth International Conference on Learning Analytics and Knowledge, 103 – 112. Association for Computing Machinery. https://doi.org/10.1145/2567574.2567583.</bibtext> </blist> <blist> <bibtext> Ahn, D. 2024. "Using Interpolated Testing to Improve Online Learning of Community College and University Students." Unpublished doctoral diss., Iowa State University.</bibtext> </blist> <blist> <bibtext> Ahn, D., and J. C. K. Chan. 2023. " Does Testing Potentiate New Learning Because It Enables Learners to Use Better Strategies? " Journal of Experimental Psychology. Learning, Memory, and Cognition 50, no. 3 : 435 – 457. https://doi.org/10.1037/xlm0001233.</bibtext> </blist> <blist> <bibtext> Ahn, D., J. C. K. Chan, K. K. Szpunar, H. Gill, and Z. Assadipour. 2023. " Can Testing Enhance Learning in Distracting Online Environments? " In Society for Applied Research in Memory and Cognition (SARMAC) XIV 2023.</bibtext> </blist> <blist> <bibtext> Ashburner, M., S. Singh, L. Bianchi, and E. F. Risko. 2024. "Speeding Recorded Lectures: Effects on Test Performance, Metacognition, and Affect and the Importance of the Evaluative Context." Unpublished manuscript, Department of Psychology, University of Waterloo.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref190" type="bt">6</bibl> <bibtext> Bauer, M., M. Malchow, and C. Meinel. 2018. " Improving Access to Online Lecture Videos." In 2018 IEEE Global Engineering Education Conference (EDUCON), 1161 – 1168. IEEE. https://doi.org/10.1109/EDUCON.2018.8363361.</bibtext> </blist> <blist> <bibl id="bib7" idref="ref173" type="bt">7</bibl> <bibtext> Beege, M., N. L. Schroeder, S. Heidig, G. D. Rey, and S. Schneider. 2023. " The Instructor Presence Effect and Its Moderators in Instructional Video: A Series of Meta‐Analyses." Educational Research Review 41 : 100564. https://doi.org/10.1016/j.edurev.2023.100564.</bibtext> </blist> <blist> <bibl id="bib8" idref="ref1" type="bt">8</bibl> <bibtext> Cameron, M., T. A. Lacy, P. Siegel, et al. 2021. 2019–20 National Postsecondary Student Aid Study (NPSAS:20): First Look at the Impact of Coronavirus (COVID‐19) Pandemic on Undergraduate Student Enrollment, Housing, and Finances (NCES 2021‐456). National Center for Education Statistics. https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2021456.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref99" type="bt">9</bibl> <bibtext> Carpenter, S. K., M. M. Wilford, N. Kornell, and K. M. Mullaney. 2013. " Appearances Can Be Deceiving: Instructor Fluency Increases Perceptions of Learning Without Increasing Actual Learning." Psychonomic Bulletin & Review 20 : 1350 – 1356. https://doi.org/10.3758/s13423‐013‐0442‐z.</bibtext> </blist> <blist> <bibtext> Carpenter, S. K., A. E. Witherby, and S. K. Tauber. 2020. " On Students' (Mis)Judgments of Learning and Teaching Effectiveness." Journal of Applied Research in Memory and Cognition 9, no. 2 : 137 – 151. https://doi.org/10.1016/j.jarmac.2019.12.009.</bibtext> </blist> <blist> <bibtext> Ceci, S. J., S. Kahn, and W. M. Williams. 2023. " Exploring Gender Bias in Six Key Domains of Academic Science: An Adversarial Collaboration." Psychological Science in the Public Interest 24, no. 1 : 15 – 73. https://doi.org/10.1177/15291006231163179.</bibtext> </blist> <blist> <bibtext> Chan, J. C. K., and D. Ahn. 2023. " Unproctored Online Exams Provide Meaningful Assessment of Student Learning." Proceedings of the National Academy of Sciences (PNAS) 120, no. 31 : e2302020120. https://doi.org/10.1073/pnas.2302020120.</bibtext> </blist> <blist> <bibtext> Chan, J. C. K., C. A. Meissner, and S. D. Davis. 2018. " Retrieval Potentiates New Learning: A Theoretical and Meta‐Analytic Review." Psychological Bulletin 144, no. 11 : 1111 – 1146. https://doi.org/10.1037/bul0000166.</bibtext> </blist> <blist> <bibtext> Cialdini, R. B., M. R. Trost, and J. T. Newsom. 1995. " Preference for Consistency: The Development of a Valid Measure and the Discovery of Surprising Behavioral Implications." Journal of Personality and Social Psychology 69, no. 2 : 318 – 328. https://doi.org/10.1037/0022‐3514.69.2.318.</bibtext> </blist> <blist> <bibtext> Clark, R. C., and R. E. Mayer. 2016. E‐Learning and the Science of Instruction: Proven Guidelines for Consumers and Designers of Multimedia Learning. 4th ed. Wiley. https://doi.org/10.1002/9781119239086.</bibtext> </blist> <blist> <bibtext> Coursera. 2024. "About Coursera." https://about.coursera.org/.</bibtext> </blist> <blist> <bibtext> Cuevas, J. 2015. " Is Learning Styles‐Based Instruction Effective? A Comprehensive Analysis of Recent Research on Learning Styles." Theory and Research in Education 13, no. 3 : 308 – 333. https://doi.org/10.1177/1477878515606621.</bibtext> </blist> <blist> <bibtext> Dahl, T. I., and S. Ludvigsen. 2014. " How I See What You're Saying: The Role of Gestures in Native and Foreign Language Listening Comprehension." Modern Language Journal 98, no. 3 : 813 – 833. https://doi.org/10.1111/modl.12124.</bibtext> </blist> <blist> <bibtext> Davis, R. O. 2018. " The Impact of Pedagogical Agent Gesturing in Multimedia Learning Environments: A Meta‐Analysis." Educational Research Review 24 : 193 – 209. https://doi.org/10.1016/j.edurev.2018.05.002.</bibtext> </blist> <blist> <bibtext> Dixson, M. D. 2010. " Creating Effective Student Engagement in Online Courses: What Do Students Find Engaging? " Journal of the Scholarship of Teaching and Learning 10, no. 2 : 1 – 13. https://scholarworks.iu.edu/journals/index.php/josotl/article/view/1744.</bibtext> </blist> <blist> <bibtext> Doubleday, A. F., and L. M. Lee. 2016. " Dissecting the Voice: Health Professions Students' Perceptions of Instructor Age and Gender in an Online Environment and the Impact on Evaluations for Faculty." Anatomical Sciences Education 9, no. 6 : 537 – 544. https://doi.org/10.1002/ase.1609.</bibtext> </blist> <blist> <bibtext> Falk, A., and F. Zimmermann. 2011. "Preferences for Consistency." CEPR Discussion Paper No. DP8519. https://ssrn.com/abstract=1928435.</bibtext> </blist> <blist> <bibtext> Fallace, T. 2023. " The Long Origins of the Visual, Auditory, and Kinesthetic Learning Style Typology, 1921–2001." History of Psychology 26, no. 4 : 334 – 354. https://doi.org/10.1037/hop0000240.</bibtext> </blist> <blist> <bibtext> Faul, F., E. Erdfelder, A. G. Lang, and A. Buchner. 2007. " G*Power 3: A Flexible Statistical Power Analysis Program for the Social, Behavioral, and Biomedical Sciences." Behavior Research Methods 39 : 175 – 191. https://doi.org/10.3758/BF03193146.</bibtext> </blist> <blist> <bibtext> Finn, B. 2020. " What More Can We Learn From Teaching Evaluations? " Journal of Applied Research in Memory and Cognition 9, no. 2 : 157 – 160. https://doi.org/10.1016/j.jarmac.2020.02.002.</bibtext> </blist> <blist> <bibtext> Fraysier, K., A. Reschly, and J. Appleton. 2020. " Predicting Postsecondary Enrollment With Secondary Student Engagement Data." Journal of Psychoeducational Assessment 38, no. 7 : 882 – 899. https://doi.org/10.1177/0734282920903168.</bibtext> </blist> <blist> <bibtext> Fredricks, J. A. 2022. " The Measurement of Student Engagement: Methodological Advances and Comparison of New Self‐Report Instruments." In Handbook of Research on Student Engagement, edited by A. L. Reschly and S. L. Christenson, 2nd ed., 597 – 616. Springer. https://doi.org/10.1007/978‐3‐031‐07853‐8_29.</bibtext> </blist> <blist> <bibtext> Gray, C. C., and D. Perkins. 2019. " Utilizing Early Engagement and Machine Learning to Predict Student Outcomes." Computers & Education 131 : 22 – 32. https://doi.org/10.1016/j.compedu.2018.12.006.</bibtext> </blist> <blist> <bibtext> Groccia, J. E. 2018. " What is Student Engagement? " New Directions for Teaching and Learning 2018, no. 154 : 11 – 20. https://doi.org/10.1002/tl.20287.</bibtext> </blist> <blist> <bibtext> Guo, P. J., J. Kim, and R. Rubin. 2014. " How Video Production Affects Student Engagement: An Empirical Study of MOOC Videos." In Proceedings of the First ACM Conference on Learning at Scale, 41 – 50. Association for Computing Machinery. https://doi.org/10.1145/2556325.2566239.</bibtext> </blist> <blist> <bibtext> Harp, S. F., and R. E. Mayer. 1998. " How Seductive Details Do Their Damage: A Theory of Cognitive Interest in Science Learning." Journal of Educational Psychology 90, no. 3 : 414 – 434. https://doi.org/10.1037/0022‐0663.90.3.414.</bibtext> </blist> <blist> <bibtext> Homer, B. D., J. L. Plass, and L. Blake. 2008. " The Effects of Video on Cognitive Load and Social Presence in Multimedia‐Learning." Computers in Human Behavior 24, no. 3 : 786 – 797. https://doi.org/10.1016/j.chb.2007.02.009.</bibtext> </blist> <blist> <bibtext> Husmann, P. R., and V. D. O'Loughlin. 2019. " Another Nail in the Coffin for Learning Styles? Disparities Among Undergraduate Anatomy Students' Study Strategies, Class Performance, and Reported Vark Learning Styles." Anatomical Sciences Education 12, no. 1 : 6 – 19. https://doi.org/10.1002/ase.1777.</bibtext> </blist> <blist> <bibtext> JASP Team. 2024. "JASP (Version 0.18.3) [Computer software]."</bibtext> </blist> <blist> <bibtext> Kizilcec, R. F., J. N. Bailenson, and C. J. Gomez. 2015. " The Instructor's Face in Video Instruction: Evidence From Two Large‐Scale Field Studies." Journal of Educational Psychology 107, no. 3 : 724 – 739. https://doi.org/10.1037/edu0000013.</bibtext> </blist> <blist> <bibtext> Kizilcec, R. F., K. Papadopoulos, and L. Sritanyaratana. 2014. " Showing Face in Video Instruction: Effects on Information Retention, Visual Attention, and Affect." In Proceedings of the Annual SIGCHI Conference on Human Factors in Computing Systems, 2095 – 2102. Association for Computing Machinery. https://doi.org/10.1145/2556288.2557207.</bibtext> </blist> <blist> <bibtext> Krause, K.‐L., and H. Coates. 2008. " Students' Engagement in First‐Year University." Assessment & Evaluation in Higher Education 33, no. 5 : 493 – 505. http://10.1080/02602930701698892.</bibtext> </blist> <blist> <bibtext> Kuh, G. D. 2001. The National Survey of Student Engagement: Conceptual Framework and Overview of Psychometric Properties. Indiana University, Center for Postsecondary Research.</bibtext> </blist> <blist> <bibtext> Kuh, G. D. 2003. " What We're Learning About Student Engagement From NSSE: Benchmarks for Effective Educational Practices." Change 35, no. 2 : 24 – 32.</bibtext> </blist> <blist> <bibtext> Kuh, G. D., T. M. Cruce, R. Shoup, J. Kinzie, and R. M. Gonyea. 2008. " Unmasking the Effects of Student Engagement on First‐Year College Grades and Persistence." Journal of Higher Education 79, no. 5 : 540 – 563.</bibtext> </blist> <blist> <bibtext> Lackmann, S., P. M. Léger, P. Charland, C. Aubé, and J. Talbot. 2021. " The Influence of Video Format on Engagement and Performance in Online Learning." Brain Sciences 11, no. 2 : 1 – 21. https://doi.org/10.3390/brainsci11020128.</bibtext> </blist> <blist> <bibtext> Lagerstrom, L., P. Johanes, and U. Ponsukcharoen. 2015. " 2015 ASEE Annual Conference & Exposition." In The Myth of the Six‐Minute Rule: Student Engagement With Online Videos, 1 – 17. ASEE. https://peer.asee.org/24895.</bibtext> </blist> <blist> <bibtext> Lang, D., G. Chen, K. Mirzaei, and A. Paepcke. 2020. " Is Faster Better? A Study of Video Playback Speed." In Proceedings of the Tenth International Conference on Learning Analytics & Knowledge, 260 – 269. Association for Computing Machinery. https://doi.org/10.1145/3375462.3375466.</bibtext> </blist> <blist> <bibtext> Li, W., Z. Kuang, X. Leng, R. E. Mayer, and F. Wang. 2024. " Role of Gesturing Onscreen Instructors in Video Lectures: A Set of Three‐level Meta‐analyses on the Embodiment Effect." Educational Psychology Review 36, no. 3 : 67.</bibtext> </blist> <blist> <bibtext> Manasrah, A., M. Masoud, and Y. Jaradat. 2021. " Short Videos, or Long Videos? A Study on the Ideal Video Length in Online Learning." In 2021 International Conference on Information Technology (ICIT), 366 – 370. IEEE. https://doi.org/10.1109/ICIT52682.2021.9491115.</bibtext> </blist> <blist> <bibtext> Marcus, T., J. King, and J. Markant. 2023. Visual Cues and Peer Videos Differentially Affect College Students' Looking Behaviors and Learning During Online Lessons. Poster Presented At the Annual 64th convention of the Psychonomic Society.</bibtext> </blist> <blist> <bibtext> Matz, S. C., C. S. Bukow, H. Peters, C. Deacons, A. Dinu, and C. Stachl. 2023. " Using Machine Learning to Predict Student Retention From Socio‐Demographic Characteristics and App‐Based Engagement Metrics." Scientific Reports 13, no. 1 : 5705. https://doi.org/10.1038/s41598‐023‐32484‐w.</bibtext> </blist> <blist> <bibtext> Mayer, R. E. 2023. " Research‐Based Principles for Designing Multimedia Instruction." In In Their Own Words: What Scholars and Teachers Want You to Know About Why and How to Apply the Science of Learning in Your Academic Setting, edited by C. E. Overson, C. M. Hakala, L. L. Kordonowy, and V. A. Benassi, 143 – 157. Society for the Teaching of Psychology. https://teachpsych.org/ebooks/itow.</bibtext> </blist> <blist> <bibtext> Mayer, R. E., G. T. Dow, and S. Mayer. 2003. " Multimedia Learning in an Interactive Self‐Explaining Environment: What Works in the Design of Agent‐Based Microworlds? " Journal of Educational Psychology 95, no. 4 : 806 – 812. https://doi.org/10.1037/0022‐0663.95.4.806.</bibtext> </blist> <blist> <bibtext> McDermott, K. B. 2021. " Practicing Retrieval Facilitates Learning." Annual Review of Psychology 72 : 609 – 633. https://doi.org/10.1146/annurev‐psych‐010419‐051019.</bibtext> </blist> <blist> <bibtext> MOOC.org. 2024. "About MOOCs." https://<ulink href="http://www.mooc.org/about‐moocs">www.mooc.org/about‐moocs</ulink>.</bibtext> </blist> <blist> <bibtext> Moreno, R., R. E. Mayer, H. A. Spires, and J. C. Lester. 2001. " The Case for Social Agency in Computer‐Based Teaching: Do Students Learn More Deeply When They Interact With Animated Pedagogical Agents? " Cognition and Instruction 19, no. 2 : 177 – 213. https://doi.org/10.1207/S1532690XCI1902_02.</bibtext> </blist> <blist> <bibtext> Murphy, D. H., K. M. Hoover, K. Agadzhanyan, J. C. Kuehn, and A. D. Castel. 2022. " Learning in Double Time: The Effect of Lecture Video Speed on Immediate and Delayed Comprehension." Applied Cognitive Psychology 36, no. 1 : 69 – 82. https://doi.org/10.1002/acp.3899.</bibtext> </blist> <blist> <bibtext> Murphy, D. H., K. M. Hoover, and A. D. Castel. 2023. " The Effect of Video Playback Speed on Learning and Mind‐Wandering in Younger and Older Adults." Memory 31, no. 6 : 802 – 817. https://doi.org/10.1080/09658211.2023.2198326.</bibtext> </blist> <blist> <bibtext> Nagahama, T., and Y. Morita. 2017. " Effect Analysis of Playback Speed for Lecture Video Including Instructor Images." International Journal for Educational Media and Technology 11, no. 1 : 50 – 58. https://ijemt.org/index.php/journal/article/view/217.</bibtext> </blist> <blist> <bibtext> National Center for Education Statistics. 2023. Number and Percentage of Students Enrolled in Degree‐Granting Postsecondary Institutions, by Distance Education Participation, Location of Student, Level of Enrollment, and Control and Level of Institution: Fall 2021 and Fall 2022. National Center for Education Statistics. https://nces.ed.gov/programs/digest/d23/tables/dt23_311.15.asp.</bibtext> </blist> <blist> <bibtext> OpenAI. 2023. "GPT (Oct 6 Version) [Large Language Model]." https://chat.openai.com/chat.</bibtext> </blist> <blist> <bibtext> Ozan, O., and Y. Ozarslan. 2016. " Video Lecture Watching Behaviors of Learners in Online Courses." Educational Media International 53, no. 1 : 27 – 41. https://doi.org/10.1080/09523987.2016.1189255.</bibtext> </blist> <blist> <bibtext> Park, S., S. D. Holloway, and A. Arendtsz. 2012. " What Makes Students Engaged in Learning? A Time‐Use Study of Within‐ and Between‐Individual Predictors of Emotional Engagement in Low‐Performing High Schools." Journal of Youth and Adolescence 41 : 390 – 401. https://doi.org/10.1007/s10964‐011‐9738‐3.</bibtext> </blist> <blist> <bibtext> Pashler, H., M. McDaniel, D. Rohrer, and R. Bjork. 2008. " Learning Styles: Concepts and Evidence." Psychological Science in the Public Interest 9, no. 3 : 105 – 119. https://doi.org/10.1111/j.1539‐6053.2009.01038.x.</bibtext> </blist> <blist> <bibtext> Perloff, R. M. 2003. The Dynamics of Persuasion: Communication and Attitudes in the 21st Century. 2nd ed. Lawrence Erlbaum Associates.</bibtext> </blist> <blist> <bibtext> Pi, Z., J. Hong, and J. Yang. 2017. " Does Instructor's Image Size in Video Lectures Affect Learning Outcomes? " Journal of Computer Assisted Learning 33, no. 4 : 347 – 354. https://doi.org/10.1111/jcal.12183.</bibtext> </blist> <blist> <bibtext> Piquard‐Kipffer, A., T. Cavadini, L. Sprenger‐Charolles, and E. Gentaz. 2021. " Impact of Lip‐Reading on Speech Perception in French‐Speaking Children at Risk for Reading Failure Assessed From Age 5 to 7." L'Année Psychologique 121, no. 2 : 3 – 18.</bibtext> </blist> <blist> <bibtext> Rea, S. D., L. Wang, K. Muenks, and V. X. Yan. 2022. " Students Can (Mostly) Recognize Effective Learning, So Why Do They Not Do It? " Journal of Intelligence 10, no. 4 : 127. https://doi.org/10.3390/jintelligence10040127.</bibtext> </blist> <blist> <bibtext> Reeves, B., and C. I. Nass. 1996. How People Treat Computers, Television, and New Media Like Real People and Places. CSLI Publications.</bibtext> </blist> <blist> <bibtext> Richardson, J., A. Koehler, E. Besser, S. Caskurlu, J. Lim, and C. Mueller. 2015. " Conceptualizing and Investigating Instructor Presence in Online Learning Environments." International Review of Research in Open and Distance Learning 16, no. 3 : 256 – 297. https://doi.org/10.19173/irrodl.v16i3.2123.</bibtext> </blist> <blist> <bibtext> Risko, E. F., J. Liu, and L. Bianchi. 2024. " Speeding Lectures to Make Time for Retrieval Practice: Can We Improve the Efficiency of Interpolated Testing? " Journal of Experimental Psychology. Applied 30, no. 2 : 268 – 281. https://doi.org/10.1037/xap0000494.</bibtext> </blist> <blist> <bibtext> Rogowsky, B. A., B. M. Calhoun, and P. Tallal. 2015. " Matching Learning Style to Instructional Method: Effects on Comprehension." Journal of Educational Psychology 107, no. 1 : 64 – 78. https://doi.org/10.1037/a0037478.</bibtext> </blist> <blist> <bibtext> Rogowsky, B. A., B. M. Calhoun, and P. Tallal. 2020. " Providing Instruction Based on Students' Learning Style Preferences Does Not Improve Learning." Frontiers in Psychology 11 : 164. https://doi.org/10.3389/fpsyg.2020.00164.</bibtext> </blist> <blist> <bibtext> Schwartz, J. L., F. Berthommier, and C. Savariaux. 2004. " Seeing to Hear Better: Evidence for Early Audio‐Visual Interactions in Speech Identification." Cognition 93, no. 2 : B69 – B78. https://doi.org/10.1016/j.cognition.2004.01.006.</bibtext> </blist> <blist> <bibtext> Sepp, S., M. Wong, V. Hoogerheide, and J. C. Castro‐Alonso. 2022. " Shifting Online: 12 Tips for Online Teaching Derived From Contemporary Educational Psychology Research." Journal of Computer Assisted Learning 38, no. 5 : 1304 – 1320. https://doi.org/10.1111/jcal.12715.</bibtext> </blist> <blist> <bibtext> Sheridan, K., and M. A. Kelly. 2010. " The Indicators of Instructor Presence That Are Important to Students in Online Courses." Journal of Online Learning and Teaching 6, no. 4 : 767 – 779.</bibtext> </blist> <blist> <bibtext> Shindo, M., K. Kaga, and Y. Tanaka. 1991. " Speech Discrimination and Lip Reading in Patients With Word Deafness or Auditory Agnosia." Brain and Language 40, no. 2 : 153 – 161. https://doi.org/10.1016/0093‐934X(91)90122‐H.</bibtext> </blist> <blist> <bibtext> Simonds, B. K., K. R. Meyer, M. M. Quinlan, and S. K. Hunt. 2006. " Effects of Instructor Speech Rate on Student Affective Learning, Recall, and Perceptions of Nonverbal Immediacy, Credibility, and Clarity." Communication Research Reports 23, no. 3 : 187 – 197. https://doi.org/10.1080/08824090600796401.</bibtext> </blist> <blist> <bibtext> Sinatra, G. M., B. C. Heddy, and D. Lombardi. 2015. " The Challenges of Defining and Measuring Student Engagement in Science." Educational Psychologist 50, no. 1 : 1 – 13. https://doi.org/10.1080/00461520.2014.1002924.</bibtext> </blist> <blist> <bibtext> Skipper, J. I., H. C. Nusbaum, and S. L. Small. 2005. " Listening to Talking Faces: Motor Cortical Activation During Speech Perception." NeuroImage 25, no. 1 : 76 – 89. https://doi.org/10.1016/j.neuroimage.2004.11.006.</bibtext> </blist> <blist> <bibtext> Sondermann, C., and M. Merkt. 2023a. " What Is the Effect of Talking‐Heads in Educational Videos With Different Types of Narrated Slides? " Contemporary Educational Psychology 74 : 102207. https://doi.org/10.1016/j.cedpsych.2023.102207.</bibtext> </blist> <blist> <bibtext> Sondermann, C., and M. Merkt. 2023b. " Like It or Learn From It: Effects of Talking‐Heads in Educational Videos." Computers & Education 193 : 104675. https://doi.org/10.1016/j.compedu.2022.104675.</bibtext> </blist> <blist> <bibtext> Song, K., A. Chakraborty, M. Dawson, A. Dugan, B. Adkins, and C. Doty. 2018. " Does the Podcast Video Playback Speed Affect Comprehension for Novel Curriculum Delivery? A Randomized Trial." Western Journal of Emergency Medicine 19, no. 1 : 101 – 105. http://10.5811/westjem.2017.10.36027.</bibtext> </blist> <blist> <bibtext> Sueyoshi, A., and D. M. Hardison. 2005. " The Role of Gestures and Facial Cues in Second Language Listening Comprehension." Language Learning 55, no. 4 : 661 – 699. https://doi.org/10.1111/j.0023‐8333.2005.00320.x.</bibtext> </blist> <blist> <bibtext> Sweller, J. 2011. " Cognitive Load Theory." In The Psychology of Learning and Motivation, edited by J. P. Mestre and B. H. Ross, vol. 55, 37 – 76. Elsevier Academic Press. https://doi.org/10.1016/B978‐0‐12‐387691‐1.00002‐8.</bibtext> </blist> <blist> <bibtext> Taber, K. S. 2018. " The Use of Cronbach's Alpha When Developing and Reporting Research Instruments in Science Education." Research in Science Education 48 : 1273 – 1296. https://doi.org/10.1007/s11165‐016‐9602‐2.</bibtext> </blist> <blist> <bibtext> Tavakol, M., and R. Dennick. 2011. " Making Sense of Cronbach's Alpha." International Journal of Medical Education 2 : 53 – 55. https://doi.org/10.5116/ijme.4dfb.8dfd.</bibtext> </blist> <blist> <bibtext> Taylor, S. S., and M. Statler. 2014. " Material Matters: Increasing Emotional Engagement in Learning." Journal of Management Education 38, no. 4 : 586 – 607. https://doi.org/10.1177/1052562913489976.</bibtext> </blist> <blist> <bibtext> Tharumalingam, T., and E. F. Risko. 2025. " Altering the Playback Speed of Recorded Lectures as a Learning Technique: Examining Student Practices, Motivations, and Beliefs." Journal of Research on Technology in Education : 1 – 20. https://doi.org/10.1080/15391523.2025.2459130.</bibtext> </blist> <blist> <bibtext> Tran, S., L. J. Bianchi, and E. F. Risko. 2024. " Examining Increasing Playback Speed in Recorded Lectures on Memory, Attention, and Experience." Journal of Experimental Education. https://doi.org/10.1080/00220973.2024.2306399.</bibtext> </blist> <blist> <bibtext> van Gog, T., I. Verveer, and L. Verveer. 2014. " Learning From Video Modeling Examples: Effects of Seeing the Human Model's Face." Computers & Education 72 : 323 – 327. https://doi.org/10.1016/j.compedu.2013.12.004.</bibtext> </blist> <blist> <bibtext> van Wermeskerken, M., and T. van Gog. 2017. " Seeing the Instructor's Face and Gaze in Demonstration Video Examples Affects Attention Allocation but Not Learning." Computers & Education 113 : 98 – 107. https://doi.org/10.1016/j.compedu.2017.05.013.</bibtext> </blist> <blist> <bibtext> Wang, J., P. Antonenko, and K. Dawson. 2020. " Does Visual Attention to the Instructor in Online Video Affect Learning and Learner Perceptions? An Eye‐Tracking Analysis." Computers & Education 146 : 103779. https://doi.org/10.1016/j.compedu.2019.103779.</bibtext> </blist> <blist> <bibtext> Wang, J., and P. D. Antonenko. 2017. " Instructor Presence in Instructional Video: Effects on Visual Attention, Recall, and Perceived Learning." Computers in Human Behavior 71 : 79 – 89. https://doi.org/10.1016/j.chb.2017.01.049.</bibtext> </blist> <blist> <bibtext> Wang, M., J. A. Fredricks, F. Ye, T. L. Hofkens, and J. S. Linn. 2016. " The Math and Science Engagement Scales: Scale Development, Validation, and Psychometric Properties." Learning and Instruction 43 : 16 – 26. https://doi.org/10.1016/j.learninstruc.2016.01.008.</bibtext> </blist> <blist> <bibtext> Was, C. A., R. B. Hollis, and J. Dunlosky. 2019. " Do Students Understand the Detrimental Effects of Mind Wandering During Online Learning? " Computers in Education 135 : 113 – 122. https://doi.org/10.1016/j.compedu.2019.02.020.</bibtext> </blist> <blist> <bibtext> Wells, G. L., and P. D. Windschitl. 1999. " Stimulus Sampling and Social Psychological Experimentation." Personality and Social Psychology Bulletin 25, no. 9 : 1115 – 1125. https://doi.org/10.1177/0146167299251.</bibtext> </blist> <blist> <bibtext> Wilson, K. E., M. Martinez, C. Mills, S. D'Mello, D. Smilek, and E. F. Risko. 2018. " Instructor Presence Effect: Liking Does Not Always Lead to Learning." Computers & Education 122 : 205 – 220. https://doi.org/10.1016/j.compedu.2018.03.011.</bibtext> </blist> <blist> <bibtext> Wong, M., L. M. Marshall, H. C. Blank, and B. M. Hard. 2022. " Up Close and Personal: Examining Effects of Instructor Video Presence on Student's Sense of Connection." Scholarship of Teaching and Learning in Psychology 8, no. 4 : 374 – 392. https://doi.org/10.1037/stl0000306.</bibtext> </blist> </ref> <aug> <p>By Dahwi Ahn and Jason C. K. Chan</p> <p>Reported by Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib12" firstref="ref2"></nolink> <nolink nlid="nl2" bibid="bib56" firstref="ref3"></nolink> <nolink nlid="nl3" bibid="bib16" firstref="ref4"></nolink> <nolink nlid="nl4" bibid="bib51" firstref="ref5"></nolink> <nolink nlid="nl5" bibid="bib45" firstref="ref7"></nolink> <nolink nlid="nl6" bibid="bib58" firstref="ref8"></nolink> <nolink nlid="nl7" bibid="bib30" firstref="ref9"></nolink> <nolink nlid="nl8" bibid="bib92" firstref="ref12"></nolink> <nolink nlid="nl9" bibid="bib38" firstref="ref13"></nolink> <nolink nlid="nl10" bibid="bib39" firstref="ref14"></nolink> <nolink nlid="nl11" bibid="bib40" firstref="ref15"></nolink> <nolink nlid="nl12" bibid="bib28" firstref="ref17"></nolink> <nolink nlid="nl13" bibid="bib47" firstref="ref18"></nolink> <nolink nlid="nl14" bibid="bib25" firstref="ref19"></nolink> <nolink nlid="nl15" bibid="bib26" firstref="ref20"></nolink> <nolink nlid="nl16" bibid="bib20" firstref="ref21"></nolink> <nolink nlid="nl17" bibid="bib66" firstref="ref22"></nolink> <nolink nlid="nl18" bibid="bib72" firstref="ref23"></nolink> <nolink nlid="nl19" bibid="bib36" firstref="ref24"></nolink> <nolink nlid="nl20" bibid="bib44" firstref="ref25"></nolink> <nolink nlid="nl21" bibid="bib77" firstref="ref26"></nolink> <nolink nlid="nl22" bibid="bib94" firstref="ref27"></nolink> <nolink nlid="nl23" bibid="bib90" firstref="ref28"></nolink> <nolink nlid="nl24" bibid="bib95" firstref="ref29"></nolink> <nolink nlid="nl25" bibid="bib35" firstref="ref30"></nolink> <nolink nlid="nl26" bibid="bib53" firstref="ref32"></nolink> <nolink nlid="nl27" bibid="bib67" firstref="ref33"></nolink> <nolink nlid="nl28" bibid="bib37" firstref="ref34"></nolink> <nolink nlid="nl29" bibid="bib75" firstref="ref36"></nolink> <nolink nlid="nl30" bibid="bib27" firstref="ref37"></nolink> <nolink nlid="nl31" bibid="bib91" firstref="ref39"></nolink> <nolink nlid="nl32" bibid="bib15" firstref="ref47"></nolink> <nolink nlid="nl33" bibid="bib48" firstref="ref48"></nolink> <nolink nlid="nl34" bibid="bib65" firstref="ref49"></nolink> <nolink nlid="nl35" bibid="bib19" firstref="ref50"></nolink> <nolink nlid="nl36" bibid="bib52" firstref="ref51"></nolink> <nolink nlid="nl37" bibid="bib87" firstref="ref53"></nolink> <nolink nlid="nl38" bibid="bib88" firstref="ref55"></nolink> <nolink nlid="nl39" bibid="bib31" firstref="ref56"></nolink> <nolink nlid="nl40" bibid="bib81" firstref="ref57"></nolink> <nolink nlid="nl41" bibid="bib71" firstref="ref59"></nolink> <nolink nlid="nl42" bibid="bib46" firstref="ref62"></nolink> <nolink nlid="nl43" bibid="bib49" firstref="ref64"></nolink> <nolink nlid="nl44" bibid="bib29" firstref="ref68"></nolink> <nolink nlid="nl45" bibid="bib59" firstref="ref69"></nolink> <nolink nlid="nl46" bibid="bib84" firstref="ref70"></nolink> <nolink nlid="nl47" bibid="bib60" firstref="ref73"></nolink> <nolink nlid="nl48" bibid="bib17" firstref="ref74"></nolink> <nolink nlid="nl49" bibid="bib23" firstref="ref75"></nolink> <nolink nlid="nl50" bibid="bib33" firstref="ref76"></nolink> <nolink nlid="nl51" bibid="bib68" firstref="ref77"></nolink> <nolink nlid="nl52" bibid="bib69" firstref="ref78"></nolink> <nolink nlid="nl53" bibid="bib89" firstref="ref85"></nolink> <nolink nlid="nl54" bibid="bib78" firstref="ref87"></nolink> <nolink nlid="nl55" bibid="bib93" firstref="ref98"></nolink> <nolink nlid="nl56" bibid="bib10" firstref="ref100"></nolink> <nolink nlid="nl57" bibid="bib11" firstref="ref101"></nolink> <nolink nlid="nl58" bibid="bib21" firstref="ref102"></nolink> <nolink nlid="nl59" bibid="bib57" firstref="ref103"></nolink> <nolink nlid="nl60" bibid="bib24" firstref="ref104"></nolink> <nolink nlid="nl61" bibid="bib34" firstref="ref106"></nolink> <nolink nlid="nl62" bibid="bib82" firstref="ref107"></nolink> <nolink nlid="nl63" bibid="bib83" firstref="ref108"></nolink> <nolink nlid="nl64" bibid="bib142" firstref="ref109"></nolink> <nolink nlid="nl65" bibid="bib143" firstref="ref113"></nolink> <nolink nlid="nl66" bibid="bib73" firstref="ref117"></nolink> <nolink nlid="nl67" bibid="bib43" firstref="ref122"></nolink> <nolink nlid="nl68" bibid="bib61" firstref="ref125"></nolink> <nolink nlid="nl69" bibid="bib74" firstref="ref126"></nolink> <nolink nlid="nl70" bibid="bib54" firstref="ref127"></nolink> <nolink nlid="nl71" bibid="bib55" firstref="ref128"></nolink> <nolink nlid="nl72" bibid="bib79" firstref="ref129"></nolink> <nolink nlid="nl73" bibid="bib70" firstref="ref130"></nolink> <nolink nlid="nl74" bibid="bib76" firstref="ref131"></nolink> <nolink nlid="nl75" bibid="bib80" firstref="ref132"></nolink> <nolink nlid="nl76" bibid="bib63" firstref="ref133"></nolink> <nolink nlid="nl77" bibid="bib18" firstref="ref135"></nolink> <nolink nlid="nl78" bibid="bib421" firstref="ref137"></nolink> <nolink nlid="nl79" bibid="bib281" firstref="ref147"></nolink> <nolink nlid="nl80" bibid="bib230" firstref="ref150"></nolink> <nolink nlid="nl81" bibid="bib129" firstref="ref152"></nolink> <nolink nlid="nl82" bibid="bib41" firstref="ref155"></nolink> <nolink nlid="nl83" bibid="bib62" firstref="ref156"></nolink> <nolink nlid="nl84" bibid="bib32" firstref="ref166"></nolink> <nolink nlid="nl85" bibid="bib86" firstref="ref179"></nolink> <nolink nlid="nl86" bibid="bib85" firstref="ref180"></nolink> <nolink nlid="nl87" bibid="bib64" firstref="ref181"></nolink> <nolink nlid="nl88" bibid="bib13" firstref="ref184"></nolink> <nolink nlid="nl89" bibid="bib50" firstref="ref185"></nolink> <nolink nlid="nl90" bibid="bib42" firstref="ref191"></nolink> <nolink nlid="nl91" bibid="bib14" firstref="ref192"></nolink> <nolink nlid="nl92" bibid="bib22" firstref="ref193"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1468318
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: What Drives Student Engagement and Learning in Video Lectures? An Investigation of Instructor Visibility, Playback Speed, and Student Preferences
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Dahwi+Ahn%22">Dahwi Ahn</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-4300-6520">0000-0002-4300-6520</externalLink>)<br /><searchLink fieldCode="AR" term="%22Jason+C%2E+K%2E+Chan%22">Jason C. K. Chan</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Applied+Cognitive+Psychology%22"><i>Applied Cognitive Psychology</i></searchLink>. 2025 39(2).
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 18
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: SourceSuprt
  Label: Sponsoring Agency
  Group: SrcSuprt
  Data: National Science Foundation (NSF)
– Name: NumberContract
  Label: Contract Number
  Group: NumCntrct
  Data: 2017333
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research<br />Tests/Questionnaires
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Learner+Engagement%22">Learner Engagement</searchLink><br /><searchLink fieldCode="DE" term="%22Video+Technology%22">Video Technology</searchLink><br /><searchLink fieldCode="DE" term="%22Lecture+Method%22">Lecture Method</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Attitudes%22">Student Attitudes</searchLink><br /><searchLink fieldCode="DE" term="%22Preferences%22">Preferences</searchLink><br /><searchLink fieldCode="DE" term="%22Educational+Technology%22">Educational Technology</searchLink><br /><searchLink fieldCode="DE" term="%22Time+Factors+%28Learning%29%22">Time Factors (Learning)</searchLink><br /><searchLink fieldCode="DE" term="%22Learning+Processes%22">Learning Processes</searchLink><br /><searchLink fieldCode="DE" term="%22Electronic+Learning%22">Electronic Learning</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1002/acp.70026
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0888-4080<br />1099-0720
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: COVID-19 greatly increased the online delivery of higher education. But one limitation of online learning is that students often struggle to stay engaged while watching online lectures. We examined whether including an instructor's face in lecture videos (instructor visibility) enhances student engagement or learning. In two preregistered experiments, we found that instructor visibility in lecture videos did not affect either engagement or learning overall. However, participants reported higher engagement when they watched a video that aligned with their preference for instructor visibility. For example, participants who favored videos with the instructor visible reported greater engagement with such videos compared to those without the instructor, and vice versa. Additionally, we examined the effects of playback speed on engagement and learning. Our results suggest that speeded playing did not impact engagement but resulted in better learning efficiency. Lastly, using GPT, we explored participants' open-ended responses to understand their preference for video lectures.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2025
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1468318
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1468318
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1002/acp.70026
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 18
    Subjects:
      – SubjectFull: Learner Engagement
        Type: general
      – SubjectFull: Video Technology
        Type: general
      – SubjectFull: Lecture Method
        Type: general
      – SubjectFull: Student Attitudes
        Type: general
      – SubjectFull: Preferences
        Type: general
      – SubjectFull: Educational Technology
        Type: general
      – SubjectFull: Time Factors (Learning)
        Type: general
      – SubjectFull: Learning Processes
        Type: general
      – SubjectFull: Electronic Learning
        Type: general
    Titles:
      – TitleFull: What Drives Student Engagement and Learning in Video Lectures? An Investigation of Instructor Visibility, Playback Speed, and Student Preferences
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Dahwi Ahn
      – PersonEntity:
          Name:
            NameFull: Jason C. K. Chan
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 03
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 0888-4080
            – Type: issn-electronic
              Value: 1099-0720
          Numbering:
            – Type: volume
              Value: 39
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Applied Cognitive Psychology
              Type: main
ResultId 1