View in EDS HTML Full Text PDF Full Text

The Impact of Non-Labeled Response Categories of Rating Scales: An Example with Cross-Cultural Adaptation of Two Self-Regulation Scales for Exercise

Saved in:
Bibliographic Details
Title: The Impact of Non-Labeled Response Categories of Rating Scales: An Example with Cross-Cultural Adaptation of Two Self-Regulation Scales for Exercise
Language: English
Authors: Marcela Alves Sanseverino, Ana Carolina Raabe Abitante, Monique Cristielle Silva da Silva, Liza S. Rovniak, Wagner de Lara Machado
Source: Measurement in Physical Education and Exercise Science. 2025 29(1):1-10.
Availability: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
Peer Reviewed: Y
Page Count: 10
Publication Date: 2025
Document Type: Journal Articles
Reports - Research
Descriptors: Rating Scales, Self Management, Exercise, Test Validity, Test Reliability, Translation, Psychometrics, Likert Scales, Classification, Foreign Countries, Portuguese, Measures (Individuals), Goal Orientation, Physical Activities, Adults
Geographic Terms: Brazil
DOI: 10.1080/1091367X.2024.2378875
ISSN: 1091-367X
1532-7841
Abstract: As part of a validation study of the Exercise Planning and Scheduling (EPS), and Goal-Setting (EGS) Scales, which were translated from English to Brazilian Portuguese, we aim to: present evidence of reliability and validity for the translated scale; and, explore the effects of non-labeled response categories of rating scales. The sample comprised 446 Brazilians, 82.5% female with a mean age of 32.89 (±12.21) years. The McDonald's [omega] was 0.883 and 0.899 for EPS and EGS, respectively. Descriptive data and the Rasch Models confirmed that participants tended to endorse labeled categories on 5-point Likert-type scales. The model tested with the revised response scale presented better fit indices and lower values of residual statistics than the original one. These findings contribute psychometric evidence on a translated scale that could be used in future health promotion initiatives, and suggest a psychometric characteristic -- response category labeling -- that warrants future attention in measurement development.
Abstractor: As Provided
Entry Date: 2025
Accession Number: EJ1459068
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwEQXkLBrsnc8iDV7jllkeN9AAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDGq8YXt1CVu4Hv45igIBEICBm0j57IKlw30frfeVIBKFq3wqt_QVLUfvgMMPc3cWpxxenYr5c_yghwLy8pzDjZ49LTtBNYTi3Ym-ol3F2TP_w3i7OKoyCyRhY-L9INwdYMsm0jhAS55Ho-bMSHPSEVZLhTtlZxvxiSxhkgyd05TzqcvJ57EqXZAB1x6JtxHAdicDQTqokK65OK1ZXFb8DQRtSLE4GOoNiK_6BAM9
Text:
  Availability: 1
  Value: <anid>AN0182438450;7mm01jan.25;2025Jan28.02:47;v2.2.500</anid> <title id="AN0182438450-1">The Impact of Non-Labeled Response Categories of Rating Scales: An Example with Cross-Cultural Adaptation of Two Self-Regulation Scales for Exercise </title> <p>As part of a validation study of the Exercise Planning and Scheduling (EPS), and Goal-Setting (EGS) Scales, which were translated from English to Brazilian Portuguese, we aim to: present evidence of reliability and validity for the translated scale; and, explore the effects of non-labeled response categories of rating scales. The sample comprised 446 Brazilians, 82.5% female with a mean age of 32.89 (±12.21) years. The McDonald's ω was 0.883 and 0.899 for EPS and EGS, respectively. Descriptive data and the Rasch Models confirmed that participants tended to endorse labeled categories on 5-point Likert-type scales. The model tested with the revised response scale presented better fit indices and lower values of residual statistics than the original one. These findings contribute psychometric evidence on a translated scale that could be used in future health promotion initiatives, and suggest a psychometric characteristic – response category labeling – that warrants future attention in measurement development.</p> <p>Keywords: Psychometrics; Rasch measurement; exercise; self-regulation; Brazil</p> <hd id="AN0182438450-2">Introduction</hd> <p>Physical inactivity is increasing in many countries, which results in the increase of non-communicable diseases (World Helth Organization [WHO], [<reflink idref="bib40" id="ref1">40</reflink>]). In 2018, the World Health Organization (WHO) launched the Global Action Plan on Physical Activity (GAPPA) connected with the 2030 Agenda for Sustainable Development. The GAPPA mission is to promote the implementation of interventions at the policy, community and individual levels to ensure that all people have access to safe exercise opportunities.</p> <p>In Brazil, much of the population lacks access to recreational facilities and opportunities for leisure-time physical activities (Corseuil Giehl et al., [<reflink idref="bib8" id="ref2">8</reflink>]; de Farias Júnior et al., [<reflink idref="bib10" id="ref3">10</reflink>]; Rech et al., [<reflink idref="bib32" id="ref4">32</reflink>]), and 59.5%; (IC<subs>95%</subs> = 58,8–60,2) of adults are physically inactive in their leisure time (Oliveira et al., [<reflink idref="bib28" id="ref5">28</reflink>]). In this context, individual-level interventions to help Brazilian communities boost their exercise self-regulation skills may play a valuable role in promoting physical activity (Amorim et al., [<reflink idref="bib1" id="ref6">1</reflink>]; Da Silva et al., [<reflink idref="bib9" id="ref7">9</reflink>]). The ability to self-regulate one's behavior is key for arranging the environmental supports needed to establish regular physical activity habits as part of one's daily routine (Monge-Rojas et al., [<reflink idref="bib26" id="ref8">26</reflink>]). However, there is a lack of reliable and valid scales in the Brazilian Portuguese language for assessing self-regulation skills for physical activity.</p> <hd id="AN0182438450-3">Behavioral measurement and scale labels</hd> <p>Prior self-regulation scales and other social-cognitive measures for physical activity have typically used both labeled and non-labeled response categories (i.e., some response options on a Likert-type scale are labeled with a description, and some are not and just given a number). For example, Rovniak et al. ([<reflink idref="bib34" id="ref9">34</reflink>]) developed the widely-used Exercise Planning and Scheduling (EPS) and Exercise Goal Setting (EGS) Scales, which, similar to other self-regulation scales, use both labeled and non-labeled response categories.</p> <p>Non-labeled response categories on rating scales could affect the validity parameters of psychometric instruments. It is well established that response format, anchors, and labels of rating scales can influence response patterns such as acquiescence as well as extreme- and midpoint-bias (Cabooter et al., [<reflink idref="bib5" id="ref10">5</reflink>]; Rezende & de Medeiros, [<reflink idref="bib33" id="ref11">33</reflink>]; Steinberg & Rogers, [<reflink idref="bib37" id="ref12">37</reflink>]). Dimensionality – the underlying configuration of traits or constructs that a psychometric instrument assesses, with each attribute denoting a separate dimension – may also be influenced by the use of labeled versus non-labeled response categories. This variation affects the number of factors interpreted within the same scales, contingent upon the item response structure employed (Hou et al., [<reflink idref="bib16" id="ref13">16</reflink>]). Another effect of labeled vs. non-labeled response categories may be the disorder threshold of response categories, i.e., when the probability of category endorsement does not show a monotonic pattern in relation to ability or trait level (Elliott et al., [<reflink idref="bib12" id="ref14">12</reflink>]). This means, for example, that someone with a higher level of ability might endorse a lower response category compared to someone else with a lower level of ability, resulting in a meaningless measurement structure. Each kind of response bias or a combination of them can directly impact screening or diagnostic instruments by distorting cutoff points/thresholds as well as normative data estimation and interpretation.</p> <p>Therefore, this study aimed to apply Rasch measurement to verify the psychometric properties of the translated EPS and EGS self-regulation scales. In addition, we present the process of translation and adaptation to Brazilian Portuguese of the EPS and EGS scales. Based on mathematical models that require unidimensional scales, Rasch measurement is able to evaluate the scale as a whole and its individual items within a given participant sample. For instance, we are able to verify the functioning of the response categories according to the measurement model, if the items belong to the same linear measure of the construct, if the respondents' ability is appropriate for the scale, and if the scale is able to detect variation in participants' self-regulation skills (Smith et al., [<reflink idref="bib35" id="ref15">35</reflink>]).</p> <hd id="AN0182438450-4">Material and methods</hd> <p></p> <hd id="AN0182438450-5">Recruitment</hd> <p>Participants were recruited using social media. The research team used their personal accounts on Instagram, Facebook, and WhatsApp as well as the research group account on Instagram to recruit as many people as possible. All participants volunteered to take part in the present study. The only inclusion criterion was to be 18 years or older. No exclusion criteria were applied.</p> <hd id="AN0182438450-6">Instruments</hd> <p></p> <hd id="AN0182438450-7">Sociodemographics</hd> <p>We asked participants to self-report their age, sex, educational level (seven categories ranging from incomplete middle school to completed graduate school) and, income (five categories, the first being zero to two minimum incomes, and the highest being six or more minimum incomes). For the latest, the categories were defined in accordance with the minimum income in Brazil, which at the time of data collection was R$ 1,100.00 per month (equivalent of R$ 5.00 per hour).</p> <hd id="AN0182438450-8">International physical activity questionnaire (IPAQ)</hd> <p>The IPAQ assesses physical activity over the past seven days. We used the IPAQ short form that had already been translated and adapted to the Brazilian cultural context and language (Matsudo et al., [<reflink idref="bib25" id="ref16">25</reflink>]). The validity of this instrument in the Brazilian population was previously demonstrated, showing good reproducibility (rho = 0.71) and correlation with data from a movement sensor (rho = 0.75) (Matsudo et al., [<reflink idref="bib25" id="ref17">25</reflink>]).</p> <hd id="AN0182438450-9">The exercise planning and scheduling (EPS) and exercise goal-setting (EGS) scales</hd> <p>To assess participants' planning and goal-setting practices for organizing and managing their exercise, we used the EPS and EGS scales (Rovniak et al., [<reflink idref="bib34" id="ref18">34</reflink>]). These scales previously demonstrated good internal consistency, α<subs>(EPS)</subs> = 0.87 and α<subs>(EGS)</subs> = 0.89, and test-retest reliability (<emph>r</emph> 0.89 and 0.87, respectively) in a U.S. sample of young adults. Each of the scales consists of 10 items, and the EPS has four reverse-scored items. Participants are asked to indicate the extent to which each of the 10 items describes their planning and goal-setting for exercise. Response categories range from 1 to 5, with the following category labels: 1 ="Does not Describe," 3 = "Describes Moderately," and 5 = "Describes Completely." Categories 2 and 4 of the response scale do not have any labels (refer to appendix of Rovniak et al., [<reflink idref="bib34" id="ref19">34</reflink>]).</p> <hd id="AN0182438450-10">Procedures</hd> <p>The present study was approved by the Ethics Research Committee of the Pontifical Catholic University of Rio Grande do Sul with the registration number 47,169,921.00000.5336 (CAAE). We considered key ethical considerations for online research, including data protection protocols in accordance with Brazilian law number 13,709 (Brazil, [<reflink idref="bib4" id="ref20">4</reflink>]) and Resolution 510 for research with human beings (Brazil, [<reflink idref="bib3" id="ref21">3</reflink>]). We then created an online link using Qualtrics software to assess the general population and perform the present validation study. Participants read and agreed with the consent form prior to answering any questions. If the person accessed the link but did not consent to the study, the survey was immediately terminated.</p> <p>Along with the EGS and EPS, we included the instruments previously described that were applied at the same time. The link to participate was disclosed via the research team's social media accounts. Once provided with the study link, participants used their own phones and computers to answer the questions.</p> <p>To translate the EPS and EGS scales to the Brazilian language, we followed international guidelines for instrument adaptation (International Test Comission, [<reflink idref="bib17" id="ref22">17</reflink>]; Sousa & Rojjanasrirat, [<reflink idref="bib36" id="ref23">36</reflink>]). We first contacted the original author who agreed to assist us with the necessary procedures to adapt to the Brazilian context. We then asked two independent translators, one with knowledge regarding the psychological terminology, to perform a forward translation. We then synthesized the two translations.</p> <p>Using the synthesis, we asked three expert evaluators who had their PhDs in the area of physical exercise and movement science to rate the instructions and the scale items with respect to their linguistic clarity and practical pertinence using a 1–5 point Likert-type scale, with an acceptable mean score of 4 points for retention (Pasquali, [<reflink idref="bib29" id="ref24">29</reflink>]). The mean score of linguistic clarity was 4.50 for the EGS and 4.73 for the EPS, with a range of 3.5 to 5 for the EGS and 3 to 5 for the EPS. The practical pertinence ranged from 4 to 5 with mean equal to 4.64, and 3 to 5 with a mean of 4.77 respectively. We also calculated Finn's coefficient for each item to assess the degree of interrater reliability among the expert evaluators. The mean of Finn's coefficient for language clarity was 0.71 and 0.82, and 0.82 and 0.89 for practical pertinence for the EGS and EPS, respectively, indicating good to excellent interrater reliability for the revised scales (Gödert et al., [<reflink idref="bib14" id="ref25">14</reflink>]). The expert evaluators could also comment on each scale item to suggest adjustments. All comments were taken into consideration and the necessary adjustments were made. Prior to testing the translated scales with the target population, we performed backtranslations with a third, independent translator to confirm concordance with the original English-language scales, which were reviewed and approved by the first author.</p> <hd id="AN0182438450-11">Data analyses</hd> <p>The first step was to analyze the agreement among evaluators with the Finn's Coefficient using the <emph>irr package</emph> (Gamer et al., [<reflink idref="bib13" id="ref26">13</reflink>]) in the software R version 4.0.2. Using JASP software version 0.16.4 (Han & Dawson, [<reflink idref="bib15" id="ref27">15</reflink>]), we performed descriptive analyses of the means, standard deviations, medians, and histograms of the scale items for both the EGS and EPS scales. We performed a <emph>t</emph>-test to compare the data obtained in the present study with the data from the original study regarding EPS and EGS scales. The Rasch model analysis for categorical data was performed with software WINSTEPS 3.72.2 (Linacre, [<reflink idref="bib21" id="ref28">21</reflink>]) for the original data.</p> <p>Rasch ([<reflink idref="bib31" id="ref29">31</reflink>]) model was originally designed for modeling educational assessments and nowadays represents a specific family of statistical approaches to model responses from questionnaires to experimental tasks. The Rasch models differ from other approaches, like Item Response Theory (IRT) Models, due to the estimation of only the difficulty (or location) parameter of items and the ability or latent trait (theta) parameter of persons. IRT models usually estimates an additional parameter of items' discrimination (or slope), which is highly sensitive to sample' levels of ability (<emph>i.e</emph>. latent trait) and iterates with a difficulty parameter (B. D. Wright, [<reflink idref="bib41" id="ref30">41</reflink>]). Thus, Rasch models lead to a more generalizable and sample-free estimation of items' and persons' parameters. Since the original model proposed by Greorg Rasch is designed for dichotomous responses, we chose to use the Rating Scale Model (Andrich, [<reflink idref="bib2" id="ref31">2</reflink>]), which is recommended for modeling polytomous scales such as Likert-type scales (<reflink idref="bib1" id="ref32">1</reflink>). The model assumes that each item has the same response structure, i.e., the same number of response categories. The model is given by:</p> <p>(<reflink idref="bib1" id="ref33">1</reflink>)</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi mathvariant="italic">P</mi><mfenced open="(" close=")"><mrow><mrow><mfrac><mi>X</mi><mi>t</mi></mfrac></mrow></mrow></mfenced><mo>=</mo><mi mathvariant="italic">δj</mi><mo>+</mo><mi mathvariant="italic">γk</mi><mo>+</mo><mi mathvariant="italic">ti</mi></math> </ephtml> </p> <p>Where <emph>δ</emph><subs><emph>j</emph></subs> is the location parameter for item <emph>j</emph>, plus a threshold parameter <emph>γ</emph><subs><emph>k</emph></subs> for each response option <emph>k</emph> to item <emph>j</emph>, plus a latent <emph>t</emph> parameter for subject <subs><emph>i</emph></subs>. This model permits estimation of the probability response function for each category of the rating scale (i.e., Likert-type scale options). However, in some cases, the structure of category options does not behave as an additive monotonic model, i.e., categories do not ascend according to ability/trait level, thereby forming overlapped probability curves. As this was the case with our data, a potential remedy is to modify the structure of the category responses, collapsing some of them into a more empirically guided scoring. This can lead to a different response structure across items from the same latent continuum, demanding a more specific model (<reflink idref="bib2" id="ref34">2</reflink>), such as the Partial Credit Model (Masters, [<reflink idref="bib24" id="ref35">24</reflink>]). This model is designed to give a partial score (or "credit") for each category <emph>k</emph> of item <emph>j</emph>, as follows:</p> <p>(<reflink idref="bib2" id="ref36">2</reflink>)</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi mathvariant="italic">P</mi><mfenced open="(" close=")"><mrow><mrow><mfrac><mi>X</mi><mi>t</mi></mfrac></mrow></mrow></mfenced><mo>=</mo><mi mathvariant="italic">bjk</mi><mo>+</mo><mi mathvariant="italic">ti</mi></math> </ephtml> </p> <p>Where <emph>b</emph> is a difficulty parameter for each category <emph>k</emph> of item <emph>j</emph>. This approach was useful to our study, because after recoding, the number of response categories was not equal for all items of the EPS scale. As for the EGS, it was possible to employ the original model by Andrich ([<reflink idref="bib2" id="ref37">2</reflink>]) (Equation 1).</p> <p>Finally, we computed Pearson correlations between the EGS, the EPS, and the total minutes of physical activity on the IPAQ to assess the concurrent validity of the translated EPS and EGS scales. In order to assess the reliability in a way that is comparable to other studies, we used JASP to compute McDonald's omega (ω) and Chronback's alpha (α) with the recoded scale.</p> <hd id="AN0182438450-12">Results</hd> <p>The sample of 587 participants had an average age of 32.81 (±12.15; 18–90), and most participants were female (<emph>n</emph> = 375, 82.5% of valid responses). Regarding personal income, the plurality of participants (<reflink idref="bib136" id="ref38">136</reflink>, 31.1% of valid responses) declared receiving up to two minimum income (up to approximately R$ 2,200 per month), and 110 (25.1% of valid responses) declared receiving more than 6 minimum income (R$ 6,600 or more). We recruited participants in all Brazilian regions; 259 (58.7% of valid responses) participants were from the South, 91 (20.6% of valid responses) participants were from Southeast, 34 (7.7% of valid responses) participants were from the Northeast, 31 (7.0% of valid responses) participants were from the North, and 22 participants (5.0% of valid responses) were from the Center-West. Based on the IPAQ, 144 (68,6% of valid responses) of the participants performed more than 150 min of physical exercise per week.</p> <p>The mean and standard deviation of the total scores of EGS and EPS scales were 2.13 (±1.00) and 2.79 (±1.04) with median scores of 1.85 and 2.7, respectively. In the original study, Rovniak et al. ([<reflink idref="bib34" id="ref39">34</reflink>]) presented means and SD values equals to 2.84 (±0.90) and 2.68 (±0.89), respectively. Comparing the means of the present study with the original study, we found a significant difference for EGS (test statistic <emph>t</emph> = −9.64; <emph>p</emph> <.001; effect size d = 0.74). For the EPS means, no differences were found (test statistic <emph>t</emph> = 1.46; <emph>p</emph> =.15; effect size d = 0.11).</p> <p>The descriptive analysis of both the EPS and EGS scale items indicated that the participants tended not to endorse the categories that were not labeled. We chose one item of each scale to illustrate this in Figure 1, but all items presented the same response pattern. For EPS, Figure 1 illustrates item number 3, which is a reversed item, stating, "Finding time for exercise is difficult for me," and for the EGS item 2, which states, "I usually have more than one major exercise goal."</p> <p>Graph: Figure 1. Histogram examples of both scales.</p> <p>Using Rasch Model, we were able to verify that there was a lower response rate for categories 1 and 3 of the Likert-type responses categories. Based on the rating scale model, we recoded the responses of the participants into 3 categories instead of 5. Category 0, which is labeled "Does not Describe," contains the responses of what originally were 0 and 1. Category 1 corresponds to "Describes Moderately" and contains the responses from the original category 2. Finally, category 2 corresponds to "Describes Completely" and contains the responses that were originally 3 and 4.</p> <p>Figure 2 illustrates the summary of all the items of the probability of the category response scale to be endorsed in relation to one's self-regulatory skills. For instance, a person with low self-regulatory skills is more likely to respond "Does not Describe" (0) throughout the goal-setting scale items in comparison to someone with higher self-regulatory skills. The original response scale (Figure 2) of both the EPS scales demonstrates that categories 1 and 3 (the non-labeled categories) present a low probability of endorsement compared to other labeled categories in the scale continuum. Hence the need to recode the response scale. The recoded response scale (Figure 2) shows that three response categories are sufficient to cover the different levels of self-regulatory skills for both exercise goal setting and planning/scheduling.</p> <p>Graph: Figure 2. Category probabilities: structure measures at intersections.</p> <p>Item 10 of the EPS scale ("I write my planned activity sessions in an appointment book or calendar") was an exception to the three-category modeling approach because the index of infit and outfit suggested a misfit, even after recoding into 3 categories. Therefore, for EPS 10, the item was dichotomized. The "Does not describe" included responses from previous categories of 0 and 1 and "Describes" from 2, 3 and 4.</p> <p>Table 1 presents the items statistics before and after the recoding of the responses. Infit MNSQ and outfit MNSQ measures assess the fit of data to the model. Values of these measures should be between to be considered satisfactory 0.6 to 1.4 (Smith et al., [<reflink idref="bib35" id="ref40">35</reflink>]). Overall, if comparing infit and outfit measures of the original and recoded scale, we observed that values improved. Item 10 of EPS scale continued to be above the expected range.</p> <p>Table 1. Item statistics.</p> <p> <ephtml> <table><thead><tr><td /><td>Model with original categories</td><td>Model with recoded categories</td></tr><tr><td>Item</td><td>Measure</td><td>Infit</td><td>Outfit</td><td>Measure</td><td>Infit</td><td>Outfit</td></tr></thead><tbody><tr><td>EPS 1</td><td>−0.54</td><td>0.72</td><td>0.75</td><td>−0.96</td><td>0.75</td><td>0.78</td></tr><tr><td>EPS 2</td><td>−0.57</td><td>0.80</td><td>0.83</td><td>−1.01</td><td>0.86</td><td>0.87</td></tr><tr><td>EPS 3</td><td>−0.60</td><td>0.72</td><td>0.67</td><td>−1.06</td><td>0.72</td><td>0.61</td></tr><tr><td>EPS 4</td><td>0.63</td><td>0.71</td><td>0.68</td><td>1.25</td><td>0.72</td><td>0.76</td></tr><tr><td>EPS 5</td><td>−0.25</td><td>1.33</td><td>1.69</td><td>−0.47</td><td>1.43</td><td>1.80</td></tr><tr><td>EPS 6</td><td>0.03</td><td>0.83</td><td>0.77</td><td>0.13</td><td>0.87</td><td>0.80</td></tr><tr><td>EPS 7</td><td>0.18</td><td>0.96</td><td>1.06</td><td>0.38</td><td>1.09</td><td>1.10</td></tr><tr><td>EPS 8</td><td>0.87</td><td>0.75</td><td>0.60</td><td>1.74</td><td>0.73</td><td>0.54</td></tr><tr><td>EPS 9</td><td>−0.36</td><td>1.11</td><td>1.22</td><td>−0.71</td><td>1.19</td><td>1.35</td></tr><tr><td>EPS 10</td><td>0.61</td><td>2.14</td><td>2.50</td><td>0.71</td><td>1.53</td><td>2.06</td></tr><tr><td>EGS 1</td><td>−0.49</td><td>0.87</td><td>0.77</td><td>−0.78</td><td>0.90</td><td>0.92</td></tr><tr><td>EGS 2</td><td>−0.59</td><td>1.20</td><td>0.78</td><td>−0.98</td><td>1.15</td><td>1.22</td></tr><tr><td>EGS 3</td><td>0.61</td><td>1.17</td><td>0.80</td><td>0.83</td><td>1.08</td><td>0.87</td></tr><tr><td>EGS 4</td><td>−0.64</td><td>0.89</td><td>0.78</td><td>−1.08</td><td>0.86</td><td>0.85</td></tr><tr><td>EGS 5</td><td>0.44</td><td>1.23</td><td>0.85</td><td>0.59</td><td>1.09</td><td>0.94</td></tr><tr><td>EGS 6</td><td>0.26</td><td>1.18</td><td>0.88</td><td>0.32</td><td>1.02</td><td>0.89</td></tr><tr><td>EGS 7</td><td>0.62</td><td>0.97</td><td>0.67</td><td>0.87</td><td>0.93</td><td>0.73</td></tr><tr><td>EGS 8</td><td>−0.72</td><td>0.85</td><td>1.19</td><td>−0.43</td><td>0.98</td><td>1.00</td></tr><tr><td>EGS 9</td><td>−0.01</td><td>0.90</td><td>0.77</td><td>−0.08</td><td>0.87</td><td>0.89</td></tr><tr><td>EGS 10</td><td>0.52</td><td>1.45</td><td>1.16</td><td>0.74</td><td>1.30</td><td>1.34</td></tr></tbody></table> </ephtml> </p> <p>Table 2 presents the model fit indeces that resulted from the Rasch analysis with the original and the recoded responses. Overall, these indices evaluate the quality and reliability of the measurement, which presented satisfactory values. It is important to note that the residuals indices decrease significantly, demonstrating that the recoded scales are a better fit than the original, since large residuals indicate potential areas of concern. Further details on these indices and their interpretation are provided in the Discussion section.</p> <p>Table 2. Model fit indices.</p> <p> <ephtml> <table><thead><tr><td /><td>EPS</td><td>EGS</td></tr><tr><td /><td>Original</td><td>Recoded</td><td>Original</td><td>Recoded</td></tr><tr><td>Log-Likelihood Chi square (p-value)</td><td>6295.19 (<0.001)</td><td>4092.82 (<0.001)</td><td>5550.25 (<0.001)</td><td>3333.17 (<0.001)</td></tr></thead><tbody><tr><td>Df</td><td>2619</td><td>2594</td><td>2675</td><td>2048</td></tr><tr><td>Person reliability</td><td>0.79</td><td>0.82</td><td>0.83</td><td>0.71</td></tr><tr><td>Person separation</td><td>1.95</td><td>2.14</td><td>2.19</td><td>1.58</td></tr><tr><td>Item reliability</td><td>0.99</td><td>0.99</td><td>0.98</td><td>0.97</td></tr><tr><td>Global Root-Mean-Square Residual</td><td>1.072</td><td>0.547</td><td>0.886</td><td>0.561</td></tr><tr><td>Average Root-Mean-Square Residual of response categories</td><td>0.999</td><td>0.531</td><td>0.969</td><td>0.588</td></tr></tbody></table> </ephtml> </p> <p>The reliability values of EPS recoded scale were McDonald's ω, 0.883 95% confidence interval [0.863, 0.900], and Cronbach's α, 0.874 [0.859, 0.888]. For the EGS scale, McDonald's ω was 0.899 [0.882, 0.916], and Cronbach's α, 0.895 [0.882, 0.907]. Comparing with the orginal study by Rovniak et al. ([<reflink idref="bib34" id="ref41">34</reflink>]), our results presented similar values to the previous study were α<subs>(EPS)</subs> = 0.87; α<subs>(EGS)</subs> = 0.89.</p> <p>Finally, the correlation between the EGS and EPS scores was <emph>r</emph> = 0.530 (<emph>p</emph> <.001). The total IPAQ-reported physical activity (which assesses moderate to vigorous exercise intensity) exhibited a positive correlation of <emph>r</emph> = 0.535 (<emph>p</emph> <.001) with the EPS score, and <emph>r</emph> = 0.426 (<emph>p</emph> <.001) with the EGS score, supporting concurrent validity of the translated Brazilian Portuguese scales.</p> <hd id="AN0182438450-13">Discussion</hd> <p>The present study not only translates and adapts the EGS and EPS scales successfully, but also demonstrates the importance of labeling the response scale, so participants have a clear understanding of how to correctly respond to each item. The issue of how research participants interpret the response scale of psychometric instruments has recently been brought back to light (Cabooter et al., [<reflink idref="bib5" id="ref42">5</reflink>]; Podsakoff et al., [<reflink idref="bib30" id="ref43">30</reflink>]). When researchers develop and adapt scales there are guidelines to help with psychometric decision-making, but there are only a few studies that identify the problems that can emerge from misinterpreting labeled vs. non-labeled scale categories (Cabooter et al., [<reflink idref="bib5" id="ref44">5</reflink>]; Kutscher & Eid, [<reflink idref="bib18" id="ref45">18</reflink>]) and that suggest potential remedies (Chen et al., [<reflink idref="bib7" id="ref46">7</reflink>]; Leventhal et al., [<reflink idref="bib20" id="ref47">20</reflink>]; Y. Zhang & Wang, [<reflink idref="bib43" id="ref48">43</reflink>]). In our study, we were able to present evidence that the labeling of response categories in a Likert-type scale impacts respondents' answers, and that addressing this labeling could improve the ability to detect variation in participants' self-regulatory skills.</p> <p>Another important contribution to the field was that we were able to present satisfactory evidence of the validity of the translated EGS and EPS instruments. After the recoding, the fit indices remained the same or improved in terms of adequacy levels, which supports the need to take great care when labeling the response scale. For the EGS scale, after recoding, the infit values improved, and EGS item 10 was within the expected range, indicating that the items are all contributing to the same construct measurement. Thus, it appears that all EGS items are contributing to the assessment of the goal-setting construct.</p> <p>Looking at the model indices, the log-likelihood chi-square, degrees of freedom, the global and category RMSR all decreased, indicating that the model improved. On the other hand, person separation, person reliability, and item reliability decreased, when no difference would be expected. The item reliability is not worrisome as the reduction was small and it reflects infit values, while in Table 2, it is possible to note that the values for each item improved.</p> <p>The person reliability and separation values are mathematically equivalent (B. Wright & Stone, [<reflink idref="bib42" id="ref49">42</reflink>]), so it is expected that both change proportionally. The person fit indices "measure the extent to which a person's pattern of responses to the items corresponds to that predicted by the model" (Smith et al., [<reflink idref="bib35" id="ref50">35</reflink>], p. 191). So, the recoded version of EGS may have highlighted the difficulty of this ability in our sample (Smith et al., [<reflink idref="bib35" id="ref51">35</reflink>]). Taking into consideration that the score was significantly lower in our sample relative to data presented by Rovniak et al. ([<reflink idref="bib34" id="ref52">34</reflink>]), it is likely that the goals measured by the EGS represent a difficult skill for our sample. Future research should further investigate the validity of evidence in different samples using the three categories responses.</p> <p>One advantage of Rasch analysis is that they calculate the difficulty of each item. This metric indicates how much ability a person must have in order to endorse the items. The midpoint of the difficulty scale is 0 logits, which means one has 50% of the measured ability; lower values (negatives) represent easier items and higher values (positives), that one must present a greater ability to agree with the item (López, [<reflink idref="bib22" id="ref53">22</reflink>]).</p> <p>Even though the EGS scale appears to be difficult as a whole, it is possible to divide the items into blocks according to their measure of difficulty. The first block comprises five items with a difficulty less than 50% of the ability one should have in order to endorse them, these being items 8, 4, 2, 1, and 9. Interestingly, Elavsky et al. ([<reflink idref="bib11" id="ref54">11</reflink>]) evaluated the construct of EPS and EGS scales, and, in their paper, suggested excluding items 1, 4, 8, and 9 from the EGS scale. According to the authors, the contents of these items do not represent the best practices for exercise goal setting because they express the outcome or the frequency of goal setting (Elavsky et al., [<reflink idref="bib11" id="ref55">11</reflink>]). It is important to mention that our results do not indicate the need to exclude such items, even though they indicate that a low level of ability is enough to endorse the items.</p> <p>When we look closely to blocks of difficulty form by the items of EGS, item 2 is in this first block and states, "I usually have more than one major exercise goal," which might indicate that the first aspect of goal setting ability is to stipulate the main goals to be pursued. Next is item 6, which is about monitoring one's behavior in regard to the goals that were set. Items 5 and 10 form another block demonstrating that one has to present more ability to break a major goal into smaller ones and to make them public. Finally, to address items 3 and 7, one has to develop steps and to set deadlines to achieve their goals.</p> <p>Understanding goal setting ability is important for developing effective individual-level interventions. However, recent reviews of goal setting theory (GST) have found that, so far, goal setting strategies for physical activity do not have the expected efficacy (Kwasnicka et al., [<reflink idref="bib19" id="ref56">19</reflink>]; Swann & Rosenbaum, [<reflink idref="bib38" id="ref57">38</reflink>]; Swann et al., [<reflink idref="bib39" id="ref58">39</reflink>]). It is possible that the data presented in the present paper can provide insight for adjusting and creating interventions based on goal setting ability by beginning at the first block of items which stipulates the major goals and goes on to setting smaller, more specific goals, and finally to setting deadlines to achieve them.</p> <p>In what specifically pertains to the EPS scale, a global improvement in all Rasch analysis indices was observed. The infit measure of the items indicates whether the item contributes to the unidimensional construct, and good fit values range from 0.6 to 1.4 (Smith et al., [<reflink idref="bib35" id="ref59">35</reflink>]). Table 1 shows that after recoding, the infit values were normalized into this range, with the exception of item 10, which was slightly higher than the cutoff values. This result indicates an improvement regarding the number of categories, which is supported by the increase in the personal reliability and personal separation indices (López, [<reflink idref="bib22" id="ref60">22</reflink>]). Hence, it is possible to state that the EPS scale items contribute to the measurement of the planning and scheduling of exercise and are able to measure and differentiate people's response patterns in accordance with the theoretical model (Smith et al., [<reflink idref="bib35" id="ref61">35</reflink>]).</p> <p>We observed that the infit value of EPS 10 declined significantly after recoding, which indicates the problem is not the recoding. An infit slightly higher than the cutoff values probably indicates that this item in our sample is not contributing to the exercise planning and scheduling construct. To the best of our knowledge, this is the first study to report misfit values for EPS item 10. The item should not yet be excluded because of that, but further evaluated in future studies with the new number of response categories, especially in Brazilian samples. Item 10 states, "I write my planned activity sessions in an appointment book or calendar." Usually, in Brazil, physical exercise is prescribed by a professional who accompanies their client, while other people enjoy gathering to play a sport or just going out for a run or walk (Magno et al., [<reflink idref="bib23" id="ref62">23</reflink>]); hence, it is not necessary to write down or plan their physical activity sessions.</p> <p>Dividing the EPS items in three blocks by their difficulty, the first block refers to prioritizing physical exercise and comprises item 3, 2, and 1, which are all reversed items about not having enough time to exercise. This division is supported by the literature that states that planning physical exercise has two steps, the first being to prioritize it (Elavsky et al., [<reflink idref="bib11" id="ref63">11</reflink>]). According to the authors, the second step is planning and scheduling exercise, which appears in our second block. Items 9, 5, 6, and 7 refer to planning and scheduling exercise weekly, though it is not yet a top priority. The last block comprises items 10, 4, and 8 and states that exercise is the first thing to be scheduled and all others are adjusted around it.</p> <p>The work of Elavsky et al. ([<reflink idref="bib11" id="ref64">11</reflink>]) suggested the EPS should be divided into two subscales according to their content. Taking both results together, it is possible that our data supports it being important to prioritize first and then to plan and schedule exercise. One difference is that, in our difficulty order, item 7 is "in the middle" of planning and scheduling. A possible reason for that is the content of item 7, which implies a person exercises but reduces it to take care of other, more highly-prioritized tasks, while for the first three items of the scale a person does not necessarily exercise as it is not possible to find the time at all. Furthermore, items 10, 4, and 8 are for those who for whom exercise is not only the highest priority but also the center of their schedule.</p> <p>There is evidence that the action plan is more correlated with the behavior itself than goal setting ability, as goal setting is more linked with forming the intention to exercise (Cao et al., [<reflink idref="bib6" id="ref65">6</reflink>]; Elavsky et al., [<reflink idref="bib11" id="ref66">11</reflink>]; Hou et al., [<reflink idref="bib16" id="ref67">16</reflink>]). Our findings corroborate that the EPS score presents a higher correlation with total time of physical activity than the EGS score.</p> <p>These results are also insightful for developing interventions that aim to help inactive people to increase their level of physical activity by creating an action plan. It is important to note that both abilities (EGS and EPS) are probably connected because once one breaks a major goal into smaller ones, they start to develop an action plan to implement the behavioral changes. O'Donnell et al. ([<reflink idref="bib27" id="ref68">27</reflink>]), in their program for patients with type-2 diabetes, presented the goal-setting form they encourage patients to fill out. The last question is "When will I revisit this plan?" (O'Donnell et al., [<reflink idref="bib27" id="ref69">27</reflink>], p. 2126).</p> <hd id="AN0182438450-14">Limitations and final considerations</hd> <p>One of the main limitations of the present study was not having applied the scales with three labeled response categories to another sample in order to verify whether the results with the recoded items would be sustained. In addition, to better evaluate all properties of the scales, future studies may benefit from using a more heterogeneous sample, including people with different levels of self-regulatory skills, especially according to the EGS scale. Future studies should continue to evaluate the psychometric characteristics of the EGS and EPS scales, with special attention to EPS item 10 for Brazilian samples. We also encourage researchers to use Rasch analysis and qualitative methodology (e.g., cognitive interviews) to further evaluate scale response categories. Suggestions for future studies that assess self-regulation in Brazilian communities are presented in Table 3. The present study facilitates future assessment of self-regulation among Brazilian adults by clarifying the importance of evaluating scale response categories, while contributing translated versions of the EPS and EGS scales that could be used to further promote exercise self-regulation in Brazilian communities. The analysis also allowed us to point out best practices for implementing the training of goal-setting and planning skills in interventions, while reinforcing the quality of EGS and EPS scales for measuring such abilities.</p> <p>Table 3. Recommendations for future exercise self-regulation studies in Brazil.</p> <p> <ephtml> <table><thead><tr><td>Both Scales</td><td>EGS Scale</td><td>EPS Scale</td></tr></thead><tbody><tr><td><list list-type="Bullet"><list-item><p>Study the three categories on the response rating scales;</p></list-item><list-item><p>Apply cognitive interview methodology in a heterogeneous sample;</p></list-item><list-item><p>Use Rasch model to check response categories.</p></list-item></list></td><td><list list-type="Bullet"><list-item><p>Include diverse recruitment strategies and populations to reach a heterogeneous sample regarding goal-setting ability.</p></list-item></list></td><td><list list-type="Bullet"><list-item><p>Check if item 10 ("I write my planned activity sessions in an appointment book or calendar") is meaningful for the Brazilian population;</p></list-item><list-item><p>Apply item 10 with both three and two responses categories.</p></list-item></list></td></tr></tbody></table> </ephtml> </p> <hd id="AN0182438450-15">Disclosure statement</hd> <p>No potential conflict of interest was reported by the author(s).</p> <ref id="AN0182438450-16"> <title> References </title> <blist> <bibl id="bib1" idref="ref6" type="bt">1</bibl> <bibtext> Amorim, T. C., Azevedo, M. R., & Hallal, P. C. (2010). Physical activity levels according to physical and social environmental factors in a sample of adults living in South Brazil. Journal of Physical Activity & Health, 7 (s2), S204 – S212. https://doi.org/10.1123/jpah.7.s2.s204</bibtext> </blist> <blist> <bibl id="bib2" idref="ref31" type="bt">2</bibl> <bibtext> Andrich, D. (1978). A rating formulation for ordered response categories. Psychometrika, 43 (4), 561 – 573. https://doi.org/10.1007/BF02293814</bibtext> </blist> <blist> <bibl id="bib3" idref="ref21" type="bt">3</bibl> <bibtext> Brazil. (2016, April 07). Resulotion n 510. National Health Council. https://conselho.saude.gov.br/resolucoes/2016/Reso510.pdf</bibtext> </blist> <blist> <bibl id="bib4" idref="ref20" type="bt">4</bibl> <bibtext> Brazil. (2018, August). Law nº 13709. Civil code. https://<ulink href="http://www.planalto.gov.br/ccivil%5f03/%5fato2015-2018/2018/lei/L13709compilado.htm">www.planalto.gov.br/ccivil%5f03/%5fato2015-2018/2018/lei/L13709compilado.htm</ulink></bibtext> </blist> <blist> <bibl id="bib5" idref="ref10" type="bt">5</bibl> <bibtext> Cabooter, E., Weijters, B., Geuens, M., & Vermeir, I. (2016). Scale format effects on response option interpretation and use. Journal of Business Research, 69 (7), 2574 – 2584. https://doi.org/10.1016/j.jbusres.2015.10.138</bibtext> </blist> <blist> <bibl id="bib6" idref="ref65" type="bt">6</bibl> <bibtext> Cao, D. S., Schüz, N., Xie, G. R., & Lippke, S. (2013). Planning skills moderate the intention-planning cognitions-behaviour relation: A longitudinal study on physical activity in Chinese adolescents. Research in Sports Medicine, 21 (1), 12 – 23. https://doi.org/10.1080/15438627.2012.738441</bibtext> </blist> <blist> <bibl id="bib7" idref="ref46" type="bt">7</bibl> <bibtext> Chen, H. F., Jin, K. Y., & Wang, W. C. (2017). Modified logistic regression approaches to eliminating the impact of response styles on DIF detection in likert-type scales. Frontiers in Psychology, 8, 1143. https://doi.org/10.3389/fpsyg.2017.01143</bibtext> </blist> <blist> <bibl id="bib8" idref="ref2" type="bt">8</bibl> <bibtext> Corseuil Giehl, M. W., Hallal, P. C., Brownson, R. C., & d'Orsi, E. (2017). Exploring associations between perceived measures of the environment and walking among Brazilian older adults. Journal of Aging & Health, 29 (1), 45 – 67. https://doi.org/10.1177/0898264315624904</bibtext> </blist> <blist> <bibl id="bib9" idref="ref7" type="bt">9</bibl> <bibtext> Da Silva, I. C. M., Knuth, A. G., Mielke, G. I., Azevedo, M. R., Gonçalves, H., & Hallal, P. C. (2014). Trends in leisure-time physical activity in a southern Brazilian city: 2003–2010. Journal of Physical Activity & Health, 11 (7), 1313 – 1317. https://doi.org/10.1123/jpah.2012-0460</bibtext> </blist> <blist> <bibtext> de Farias Júnior, J. C., da Silva Lopes, A., Mota, J., Santos, M. P., Ribeiro, J. C., & Hallal, P. C. (2011). Perception of the social and built environment and physical activity among Northeastern Brazil adolescents. Preventive Medicine, 52 (2), 114 – 119. https://doi.org/10.1016/j.ypmed.2010.12.002</bibtext> </blist> <blist> <bibtext> Elavsky, S., Doerksen, S. E., & Conroy, D. E. (2012). Identifying priorities among goals and plans: A critical psychometric reexamination of the exercise goal-setting and planning/scheduling scales. Sport, Exercise, & Performance Psychology, 1 (3), 158 – 172. https://doi.org/10.1037/a0028156</bibtext> </blist> <blist> <bibtext> Elliott, R., Fox, C. M., Beltyukova, S. A., Stone, G. E., Gunderson, J., & Zhang, X. (2006). Deconstructing therapy outcome measurement with rasch analysis of a measure of general clinical distress: The symptom checklist-90-revised. Psychological Assessment, 18 (4), 359 – 372. https://doi.org/10.1037/1040-3590.18.4.359</bibtext> </blist> <blist> <bibtext> Gamer, M., Lemon, J., & Singh, I. F. P. (2019). Package 'irr'. Various coefficients of interrater reliability and agreement. https://cran.r-project.org/web/packages/irr/index.html</bibtext> </blist> <blist> <bibtext> Gödert, H. W., Gamer, M., Rill, H. G., & Vossel, G. (2005). Statement validity assessment: Inter‐rater reliability of criteria‐based content analysis in the mock‐crime paradigm. Legal and Criminological Psychology, 10 (2), 225 – 245. https://doi.org/10.1348/135532505X52680</bibtext> </blist> <blist> <bibtext> Han, H., & Dawson, K. J. (2020). JASP (software). https://jasp-stats.org/</bibtext> </blist> <blist> <bibtext> Hou, B., Li, L., Zheng, L., Qi, Y., & Zhou, S. (2022). Linking exercise intention to exercise action: The moderating role of self-efficacy. Frontiers in Psychology, 13 (July), 1 – 10. https://doi.org/10.3389/fpsyg.2022.921285</bibtext> </blist> <blist> <bibtext> International Test Comission. (2017). The ITC guidelines for translating and adapting testes. https://<ulink href="http://www.intestcom.org/">www.intestcom.org/</ulink> (Second edition).</bibtext> </blist> <blist> <bibtext> Kutscher, T., & Eid, M. (2020). The effect of rating scale length on the occurrence of inappropriate category use for the assessment of job satisfaction: An experimental online study. Journal of Well-Being Assessment, 4 (1), 1 – 35. https://doi.org/10.1007/s41543-020-00024-2</bibtext> </blist> <blist> <bibtext> Kwasnicka, D., Ntoumanis, N., & Sniehotta, F. F. (2020). Setting performance and learning goals is useful for active and inactive individuals, if goals are personalized and flexible: Commentary on Swann. Health Psychology Review, 15 (1), 51 – 55. https://doi.org/10.1080/17437199.2020.1762107</bibtext> </blist> <blist> <bibtext> Leventhal, B. C., Gregg, N., & Ames, A. J. (2022). Accounting for response styles: Leveraging the benefits of combining response process data collection and response process analysis methods. Measurement: Interdisciplinary Research and Perspectives, 20 (3), 151 – 174. https://doi.org/10.1080/15366367.2021.1953315</bibtext> </blist> <blist> <bibtext> Linacre, M. (2009). WINSTEPS® Rasch measurement computer program. Version 3.72.2. Software. https://<ulink href="http://www.winsteps.com/index.htm">www.winsteps.com/index.htm</ulink></bibtext> </blist> <blist> <bibtext> López, A. T. (1998). Analisis de rasch para todos (1st ed.). Instituto de Evaluación e Ingeniería Avanzada.</bibtext> </blist> <blist> <bibtext> Magno, F., Ten Caten, C. S., Filho, A. R. R., Callegaro, A. M., & Dias Ferreira, A. D. C. (2020). Factors related to sports participation in brazil: An analysis based on the 2015 national household survey. International Journal of Environmental Research and Public Health, 17 (17), 1 – 17. https://doi.org/10.3390/ijerph17176011</bibtext> </blist> <blist> <bibtext> Masters, G. N. (1982). A rasch model for partial credit scoring. Psychometrika, 47 (2), 149 – 174. https://doi.org/10.1007/BF02296272</bibtext> </blist> <blist> <bibtext> Matsudo, S., Araújo, T., Matsudo, V., Andrade, D., Andrade, E., Oliveira, L. C., & Braggion, G. (2001). Questionário internacional DE atividade FÍSICA (ipaq): Estupo DE validade E reprodutibilidade no brasil. Atividade Física e Saúde, 6 (5), 5 – 18. <ulink href="http://rbafs.org.br/RBAFS/article/view/931/1222">http://rbafs.org.br/RBAFS/article/view/931/1222</ulink></bibtext> </blist> <blist> <bibtext> Monge-Rojas, R., Godinho, C. A., & Reyes Fernández, B. (2021). Modelling longitudinal directional associations between self-regulation, physical activity, and habit: Results from a cross-lagged panel model. International Journal of Behavioral Medicine, 28 (4), 466 – 478. https://doi.org/10.1007/s12529-020-09936-y</bibtext> </blist> <blist> <bibtext> O'Donnell, M., Carey, M. E., Horne, R., Alvarez-Iglesias, A., Davies, M. J., Byrne, M., & Dinneen, F. (2018). Assessing the effectiveness of a goal-setting session as part of a structured group self-management education programme for people with type 2 diabetes. Patient Education & Counseling, 101 (12), 2125 – 2133. https://doi.org/10.1016/j.pec.2018.07.009</bibtext> </blist> <blist> <bibtext> Oliveira, A. B. D., Katzmarzyk, P. T., Dantas, W. S., Benseñor, I. J. M., Goulart, A. D. C., & Ekelund, U. (2023). Profile of leisure-time physical activity and sedentary behavior in adults in Brazil: A nationwide survey, 2019. Epidemiologia e Serviços de Saúde, 32 (2), e2023168. https://doi.org/10.1590/s2237-96222023000200016</bibtext> </blist> <blist> <bibtext> Pasquali, L. (1999). Escalas psicométricas. In L. Pasquali (Eds.), Instrumentos psicológicos: manual prático de elaboração (pp. 105 – 126). LabPAM.</bibtext> </blist> <blist> <bibtext> Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88 (5), 879. https://doi.org/10.1037/0021-9010.88.5.879</bibtext> </blist> <blist> <bibtext> Rasch, G. (1960). Probabilistic models for some intelligence and attainment tests. University of Chicago Press.</bibtext> </blist> <blist> <bibtext> Rech, C. R., Reis, R. S., Hino, A. A., & Hallal, P. C. (2014). Personal, social and environmental correlates of physical activity in adults from Curitiba, Brazil. Preventive Medicine, 58, 53 – 57. https://doi.org/10.1016/j.ypmed.2013.10.023</bibtext> </blist> <blist> <bibtext> Rezende, N. A., & de Medeiros, D. D. (2022). How rating scales influence responses' reliability, extreme points, middle point and respondent's preferences. Journal of Business Research, Elsevier, 138 (C), 266 – 274. https://doi.org/10.1016/j.jbusres.2021.09.031</bibtext> </blist> <blist> <bibtext> Rovniak, L. S., Anderson, E. S., Winett, R. A., & Stephens, R. S. (2002). Social cognitive determinants of physical activity in young adults: A prospective structural equation analysis. Annals of Behavioral Medicine, 24 (2), 149 – 156. https://doi.org/10.1207/S15324796ABM2402_12</bibtext> </blist> <blist> <bibtext> Smith, E. V., Conrad, K. M., Chang, K., & Piazza, J. (2002). An introduction to Rasch measurement for scale development and person assessment. Journal of Nursing Measurement, 10 (3), 189 – 206. https://doi.org/10.1891/jnum.10.3.189.52562</bibtext> </blist> <blist> <bibtext> Sousa, V. D., & Rojjanasrirat, W. (2011). Translation, adaptation and validation of instruments or scales for use in cross-cultural health care research: A clear and user-friendly guideline. Journal of Evaluation in Clinical Practice, 17 (2), 268 – 274. https://doi.org/10.1111/J.1365-2753.2010.01434.X</bibtext> </blist> <blist> <bibtext> Steinberg, L., & Rogers, A. (2022). Changing the scale: The effect of modifying response scale labels on the measurement of personality and affect. Multivariate Behavioral Research, 57 (1), 79 – 93. https://doi.org/10.1080/00273171.2020.1807305</bibtext> </blist> <blist> <bibtext> Swann, C., & Rosenbaum, S. (2018). Do we need to reconsider best practice in goal setting for physical activity promotion? British Journal of Sports Medicine, 52 (8), 485 – 486. https://doi.org/10.1136/bjsports-2017-098186</bibtext> </blist> <blist> <bibtext> Swann, C., Rosenbaum, S., Lawrence, A., Vella, S. A., McEwan, D., & Ekkekakis, P. (2020). Updating goal-setting theory in physical activity promotion: A critical conceptual review. Health Psychology Review, 15 (1), 34 – 50. https://doi.org/10.1080/17437199.2019.1706616</bibtext> </blist> <blist> <bibtext> World Helth Organization [WHO]. (2020, November 26). Physical activity. https://<ulink href="http://www.who.int/news-room/fact-sheets/detail/physical-activity">www.who.int/news-room/fact-sheets/detail/physical-activity</ulink></bibtext> </blist> <blist> <bibtext> Wright, B. D. (1997). A history of social science measurement. Educational Measurement Issues & Practice, 45 (4), 51 – 71. https://doi.org/10.1111/j.1745-3992.1997.tb00606.x</bibtext> </blist> <blist> <bibtext> Wright, B., & Stone, M. (1999). Reliabilityand separation. In B. Wright & M. Stone (Eds.), Measurement essentials (2nd ed pp. 157 – 165). Wide Range, Inc. https://<ulink href="http://www.rasch.org/measess/met-19.pdf">www.rasch.org/measess/met-19.pdf</ulink></bibtext> </blist> <blist> <bibtext> Zhang, Y., & Wang, Y. (2020). Validity of three IRT models for measuring and controlling extreme and midpoint response styles. Frontiers in Psychology, 11, 271. https://doi.org/10.3389/fpsyg.2020.00271</bibtext> </blist> </ref> <aug> <p>By Marcela Alves Sanseverino; Ana Carolina Raabe Abitante; Monique Cristielle Silva da Silva; Liza S. Rovniak and Wagner de Lara Machado</p> <p>Reported by Author; Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib40" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib10" firstref="ref3"></nolink> <nolink nlid="nl3" bibid="bib32" firstref="ref4"></nolink> <nolink nlid="nl4" bibid="bib28" firstref="ref5"></nolink> <nolink nlid="nl5" bibid="bib26" firstref="ref8"></nolink> <nolink nlid="nl6" bibid="bib34" firstref="ref9"></nolink> <nolink nlid="nl7" bibid="bib33" firstref="ref11"></nolink> <nolink nlid="nl8" bibid="bib37" firstref="ref12"></nolink> <nolink nlid="nl9" bibid="bib16" firstref="ref13"></nolink> <nolink nlid="nl10" bibid="bib12" firstref="ref14"></nolink> <nolink nlid="nl11" bibid="bib35" firstref="ref15"></nolink> <nolink nlid="nl12" bibid="bib25" firstref="ref16"></nolink> <nolink nlid="nl13" bibid="bib17" firstref="ref22"></nolink> <nolink nlid="nl14" bibid="bib36" firstref="ref23"></nolink> <nolink nlid="nl15" bibid="bib29" firstref="ref24"></nolink> <nolink nlid="nl16" bibid="bib14" firstref="ref25"></nolink> <nolink nlid="nl17" bibid="bib13" firstref="ref26"></nolink> <nolink nlid="nl18" bibid="bib15" firstref="ref27"></nolink> <nolink nlid="nl19" bibid="bib21" firstref="ref28"></nolink> <nolink nlid="nl20" bibid="bib31" firstref="ref29"></nolink> <nolink nlid="nl21" bibid="bib41" firstref="ref30"></nolink> <nolink nlid="nl22" bibid="bib24" firstref="ref35"></nolink> <nolink nlid="nl23" bibid="bib136" firstref="ref38"></nolink> <nolink nlid="nl24" bibid="bib30" firstref="ref43"></nolink> <nolink nlid="nl25" bibid="bib18" firstref="ref45"></nolink> <nolink nlid="nl26" bibid="bib20" firstref="ref47"></nolink> <nolink nlid="nl27" bibid="bib43" firstref="ref48"></nolink> <nolink nlid="nl28" bibid="bib42" firstref="ref49"></nolink> <nolink nlid="nl29" bibid="bib22" firstref="ref53"></nolink> <nolink nlid="nl30" bibid="bib11" firstref="ref54"></nolink> <nolink nlid="nl31" bibid="bib19" firstref="ref56"></nolink> <nolink nlid="nl32" bibid="bib38" firstref="ref57"></nolink> <nolink nlid="nl33" bibid="bib39" firstref="ref58"></nolink> <nolink nlid="nl34" bibid="bib23" firstref="ref62"></nolink> <nolink nlid="nl35" bibid="bib27" firstref="ref68"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1459068
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: The Impact of Non-Labeled Response Categories of Rating Scales: An Example with Cross-Cultural Adaptation of Two Self-Regulation Scales for Exercise
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Marcela+Alves+Sanseverino%22">Marcela Alves Sanseverino</searchLink><br /><searchLink fieldCode="AR" term="%22Ana+Carolina+Raabe+Abitante%22">Ana Carolina Raabe Abitante</searchLink><br /><searchLink fieldCode="AR" term="%22Monique+Cristielle+Silva+da+Silva%22">Monique Cristielle Silva da Silva</searchLink><br /><searchLink fieldCode="AR" term="%22Liza+S%2E+Rovniak%22">Liza S. Rovniak</searchLink><br /><searchLink fieldCode="AR" term="%22Wagner+de+Lara+Machado%22">Wagner de Lara Machado</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Measurement+in+Physical+Education+and+Exercise+Science%22"><i>Measurement in Physical Education and Exercise Science</i></searchLink>. 2025 29(1):1-10.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 10
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Rating+Scales%22">Rating Scales</searchLink><br /><searchLink fieldCode="DE" term="%22Self+Management%22">Self Management</searchLink><br /><searchLink fieldCode="DE" term="%22Exercise%22">Exercise</searchLink><br /><searchLink fieldCode="DE" term="%22Test+Validity%22">Test Validity</searchLink><br /><searchLink fieldCode="DE" term="%22Test+Reliability%22">Test Reliability</searchLink><br /><searchLink fieldCode="DE" term="%22Translation%22">Translation</searchLink><br /><searchLink fieldCode="DE" term="%22Psychometrics%22">Psychometrics</searchLink><br /><searchLink fieldCode="DE" term="%22Likert+Scales%22">Likert Scales</searchLink><br /><searchLink fieldCode="DE" term="%22Classification%22">Classification</searchLink><br /><searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink><br /><searchLink fieldCode="DE" term="%22Portuguese%22">Portuguese</searchLink><br /><searchLink fieldCode="DE" term="%22Measures+%28Individuals%29%22">Measures (Individuals)</searchLink><br /><searchLink fieldCode="DE" term="%22Goal+Orientation%22">Goal Orientation</searchLink><br /><searchLink fieldCode="DE" term="%22Physical+Activities%22">Physical Activities</searchLink><br /><searchLink fieldCode="DE" term="%22Adults%22">Adults</searchLink>
– Name: Subject
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Brazil%22">Brazil</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1080/1091367X.2024.2378875
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1091-367X<br />1532-7841
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: As part of a validation study of the Exercise Planning and Scheduling (EPS), and Goal-Setting (EGS) Scales, which were translated from English to Brazilian Portuguese, we aim to: present evidence of reliability and validity for the translated scale; and, explore the effects of non-labeled response categories of rating scales. The sample comprised 446 Brazilians, 82.5% female with a mean age of 32.89 (±12.21) years. The McDonald's [omega] was 0.883 and 0.899 for EPS and EGS, respectively. Descriptive data and the Rasch Models confirmed that participants tended to endorse labeled categories on 5-point Likert-type scales. The model tested with the revised response scale presented better fit indices and lower values of residual statistics than the original one. These findings contribute psychometric evidence on a translated scale that could be used in future health promotion initiatives, and suggest a psychometric characteristic -- response category labeling -- that warrants future attention in measurement development.
– 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: EJ1459068
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1459068
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1080/1091367X.2024.2378875
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 10
        StartPage: 1
    Subjects:
      – SubjectFull: Rating Scales
        Type: general
      – SubjectFull: Self Management
        Type: general
      – SubjectFull: Exercise
        Type: general
      – SubjectFull: Test Validity
        Type: general
      – SubjectFull: Test Reliability
        Type: general
      – SubjectFull: Translation
        Type: general
      – SubjectFull: Psychometrics
        Type: general
      – SubjectFull: Likert Scales
        Type: general
      – SubjectFull: Classification
        Type: general
      – SubjectFull: Foreign Countries
        Type: general
      – SubjectFull: Portuguese
        Type: general
      – SubjectFull: Measures (Individuals)
        Type: general
      – SubjectFull: Goal Orientation
        Type: general
      – SubjectFull: Physical Activities
        Type: general
      – SubjectFull: Adults
        Type: general
      – SubjectFull: Brazil
        Type: general
    Titles:
      – TitleFull: The Impact of Non-Labeled Response Categories of Rating Scales: An Example with Cross-Cultural Adaptation of Two Self-Regulation Scales for Exercise
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Marcela Alves Sanseverino
      – PersonEntity:
          Name:
            NameFull: Ana Carolina Raabe Abitante
      – PersonEntity:
          Name:
            NameFull: Monique Cristielle Silva da Silva
      – PersonEntity:
          Name:
            NameFull: Liza S. Rovniak
      – PersonEntity:
          Name:
            NameFull: Wagner de Lara Machado
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 1091-367X
            – Type: issn-electronic
              Value: 1532-7841
          Numbering:
            – Type: volume
              Value: 29
            – Type: issue
              Value: 1
          Titles:
            – TitleFull: Measurement in Physical Education and Exercise Science
              Type: main
ResultId 1