Measurement Invariance of the CESD-R-10 Among Adolescents over the Transition from Paper to Online Administration

Measurement Invariance of the CESD-R-10 Among Adolescents over the Transition from Paper to Online Administration | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Measurement Invariance of the CESD-R-10 Among Adolescents over the Transition from Paper to Online Administration Mahmood R Gohari, Mark A. Ferro, Karen Patte, James MacKillop, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5815737/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Despite the prevalence of online surveys, research validating paper-based scale measurement properties for online use is lacking. We assessed the measurement invariance of the 10-item CESD-R-10 scale in adolescents transitioning from paper to online administration. We analyzed 2-year linked data from 2,831 Canadian secondary school students during 2019/20 and 2020/21. Using structural equation modeling, we examined measurement invariance across configural, metric, scalar, and strict stages due to the COVID-19 pandemic shift to online surveys. Baseline mean depression score was 8.2 (SD = 5.9), with 33.7% (N = 953) reporting clinically relevant symptoms. We found measurement invariance between paper and online administration. The results support measurement invariance in males, while partial scalar invariance for females when thresholds of items 6 (I felt fearful) and item 9 (I felt lonely) are freely estimated between the two administrations. Our findings demonstrate that the CESD-R-10 remains invariant even with a shift in data collection mode from paper to online surveys. This study supports the validity of conducting meaningful comparisons of depression symptoms when changing the mode of data collection in survey research among adolescent samples. Depression survey administration COVID-19 youth psychometric scale Introduction Adolescent depression is a significant public health concern. Recent research indicated that more than 20% of adolescents aged 12–17 years had experienced a major depressive episode and 36.7% reported consistent feelings of sadness or hopelessness within the past year (Bitsko et al., 2022 ). The significance of adolescent depression arises from its potential to result in immediate and long-term consequences. Such consequences include disruptions in educational attainment, impaired social relationships, and increased risk of suicide and substance use (Clayborne et al., 2019 ; Glied & Pine, 2002 ). The impact can extend into adulthood; those who experience depression during adolescence often face long-term psychosocial ramifications and reduced physical health and functioning in later life (Jonsson et al., 2011 ). Research is essential to advance our understanding and inform more effective interventions that can alleviate the burden of depression in adolescents. To accomplish these aims, it is imperative to employ robust measures for assessing adolescent depression. One of the most widely used depression scales is the Center for Epidemiologic Studies Depression Scale (CESD) (Radloff, 1977 ) and its shorter form versions (e.g., the CESD-R-10) (Andresen et al., 1994 ). These scales have been used in surveys with different modes of administration. In particular, during the pandemic many surveys had to move from traditional paper-based methods to an online mode due to physical distancing measures. This shift may influence participant responses. For instance, anonymity and privacy concerns may differ between paper and online modes, potentially influencing comfort of providing honest responses. Moreover, interactive elements in online surveys may influence how participants engage with and respond to questions compared to static paper-based questionnaires. Aside from the impact of the COVID-19 pandemic, online data collection has become a commonly used method due to its lower cost, time efficiency, enhanced accessibility, and increased flexibility, all enabled by advancements in technology. The ease of administration, diverse question types, multimedia integration, and adaptive questioning based on specific responses further contribute to the appeal of online surveys. Moreover, online surveys align with environmentally friendly practices by reducing reliance on paper, printing, and transportation. Considering adolescents’ ubiquitous use of and comfort with screen- and web-based platforms, surveys targeting this demographic could be more inclined to opt for online modes of data collection. Despite potential differences in performance between paper and online surveys, there remains a notable dearth of research validating the measurement properties of scales originally designed for paper administration when applied to online methodologies. To ensure reliable and valid conclusions, it is essential to demonstrate that measurement tools maintain the same factorial structure and assess the same construct across different administration modes. Measurement invariance is important, in particular, for longitudinal surveys that aim to examine trajectories of depression symptoms over time and have a change in data collection mode. Ensuring appropriate comparisons of a construct before and after administration mode change relies on establishing the equivalence and consistency of its meaning. Without this consistency, drawing comparisons and exploring trends may lead to invalid conclusions. Given the importance of measurement consistency when changing mode of data collection, the main objective of the current study is to evaluate the invariance of a 10-item version of the CESD across paper-based and online surveys conducted pre- and post- COVID-19 pandemic onset in a large sample of adolescents. Previous research has demonstrated that the survey method can notably impact the responses of gender groups to certain psychometric scales (e.g., happiness) (Zhang et al., 2017 ). Therefore, considering the observed gender differences in depression symptoms and their engagement in online surveys (Becker, 2022 ), a secondary purpose of this paper is to consider whether associations differ between gender groups. Methods Design and participation The longitudinal measurement invariance of the CESD-R-10 was studied across two repeated measurements among adolescents who participated in the COMPASS (Cannabis, Obesity, Mental health, Physical activity, Alcohol, Smoking, and Sedentary behaviour) study (Leatherdale et al., 2014 ). The sample included 2,831 secondary school students who provided two-year linked data in the 2019/20 (T1) and 2020/21 (T2) school years. Student data in T1 were collected using a paper-based machine-readable survey completed during one classroom period (Reel et al., 2020). As the COVID-19 pandemic was declared in Canada in March 2020 and schools were closed to in-person learning, T2 data were collected online using the Qualtrics XM survey software (Qualtrics, Provo, UT, USA). Student data were linked across the study waves utilizing an anonymous unique identification code, which was generated based on information derived from five initial questionnaire questions. Full school samples of students were invited to participate using an active-information passive-consent permission protocol. All the processes were granted ethical approval by the University of Waterloo (ORE#30118), Brock University (REB#18–099), CIUSSS de la Capitale-Nationale–Université Laval (#MP-13-2017-1264), and the school boards involved. Measures Depression was assessed using a 10-item version of the CESD (CESD-R-10) (Andresen et al., 1994 ). The CESD-R-10 uses a 4-point response format (0 = "None or less than 1 day"; 1 = "1–2 days"; 2 = "3–4 days"; 3 = "5–7 days") for participants to indicate the frequency to which they experienced different symptoms of unipolar depression in the last 7 days. Scores across the items were summed to create a total score ranging from a possible 0 to 30; higher scores indicated more and more frequent depression symptoms and a higher risk of clinically-relevant major depressive symptomatology. In this study, the internal consistency across CESD-R-10 items was 0.83 and 0.82 in the paper and online administered surveys, respectively. The validity and reliability of CESD-R-10 in adolescent populations have been documented (Haroz et al., 2014 ; Mossman et al., 2017 ), including cross-sectional measurement invariance by sex and grade in the COMPASS study using paper administration (Romano et al., 2021 ). Gender: Students were asked “are you a female or male?”, with response options including female, male, I describe my gender in a different way, and I prefer not to say. Due to the limited sample size of the last two options, responses were grouped into female, male, and gender-diverse/prefer not to say categories. Statistical analysis Longitudinal measurement invariance analyses were conducted within a structural equation modeling framework over the study period. Different levels of invariance were evaluated through successive comparisons of measurement models, progression from the least to the most restrictive model as outlined below (Widaman & Reise, 1997 ): Configural Invariance: This stage examined whether the general structure of the measurement model was the same across administrations. Therefore, a paper administration CFA model (2018-19 school year) was compared with models from the online survey (2019-20) to investigate configural invariance. Metric Invariance: This stage examined the equality of factor loadings across survey administration mode. The metric invariance aimed to ascertain if the scale items possessed the same meaning and strength of connection with the underlying depression construct during both measurements. Scalar Invariance: Scalar invariance was evaluated by imposing equal constraints on both factor loadings and item thresholds across the two time points. Scalar invariance implies that the scale maintains consistent measurement units and thresholds over time, and the relationship between the latent variable and its indicators remains unchanged despite change in the mode of survey administration. Strict Invariance: Further constrained added to the previous model by requiring equal residual variances across paper and online surveys to test the strict invariance. Measurement invariance was attained by assessing all four stages. If the model fit statistics were favorable at the initial stage of configural invariance, subsequent levels of the model were constrained. If the constrained values yielded a fit that was not inferior to the initial model, it was decided that the stage of measurement invariance has been attained. The Weighted Least Squares Mean and Variance adjusted (WLSMV) including THETA parametrization was used in estimating factor models because the response options were on Likert scale (Muthen et al., 2017 ). We employed two criteria to ensure measurement equivalence. The initial criterion involved ensuring that the model fit was satisfactory at every testing level. To assess the adequacy of model fit, we used the comparative fit index (CFI) with a threshold of ≥ 0.90 and the root mean square error of approximation (RMSEA) and standardized root mean square residuals (SRMR) with a threshold of ≤ 0.06 (Kline, 2023 ). The adequacy of the fit is considered satisfactory if the criteria are met by at least two of these fit indices. No statistically significant deterioration was anticipated when constraints were imposed to a model, assuming that invariance was accomplished. As such, the second criterion specified that any changes in fit indices (i.e., when moving from a model with fewer equality constraints on parameters to a more constrained model) should not exceed pre-established thresholds. We predetermined that, to establish measurement equivalence at a particular testing stage, at least two of the following fit indices, namely difference in fit indices between the two time points as ΔCFI, ΔRMSEA, or ΔSRMR, needed to meet this criterion. The defined cutoff values for changes in model fit indices were set at ΔCFI ≤ 0.010, ΔRMSEA ≤ 0.015, or ΔSRMR ≤ 0.030 (Chen, 2007 ). In cases where measurement invariance at a specific testing level was not achieved (signified by a notable deterioration in model fit), we examined partial measurement invariance. This examination aimed to identify constraints on relevant non-invariant parameters that could be removed to enhance the overall model fit. Subsequently, we tested partial invariance by comparing fit of the originally proposed model against the less constrained version. In addition, modifications indices and standardized residuals in the Mplus output were examined to enhance the model fit. We examined measurement invariance by gender groups. However, due to the small sample of students who self-identified as gender diverse, the stratified analysis was only conducted for males and females. Data management was performed with SAS Studio [Enterprise Edition]. Mplus version 8.4 was used for the confirmatory factor analyses and measurement invariance analyses. Results Over half of students (63.2%) identified as females and 1% were classified as gender diverse. At baseline (2018-19), students’ mean age was 14.3 years (SD:1.2, range: 12 to 17 years) and 62.0% were within grades 9–11. The mean depression score at baseline was 8.2 (SD = 5.9) and 33.7% (N = 953) of students reported clinically-relevant depression symptoms (CESD-R-10 score ≥ 10). Gender diverse students reported a higher mean depression score (11.9) than their female (9.1) and male (6.5) peers. Mean scores of scale items in the two years are presented in Table 1 . Table 1 CESD-R-10 Scale item means (SD) by administration mode in two years among COMPASS study (2018-19 and 2019-20) participants. *Item responses were coded as 0 (none or less than 1 day) to 3 (5–7 days). **Items 5 and 8 were reverse coded to be consistent with the others. Higher scores indicate more depression symptoms. Item Paper (2018-19) Online (2019-20) 1. I was bothered by things that usually don’t bother me* 0.43 (0.74) 0.50 (0.79) 2. I had trouble keeping my mind on what I was doing 0.87 (0.94) 1.04 (1.02) 3. I felt depressed 0.74 (0.98) 0.84 (1.00) 4. I felt that everything I did was an effort 0.85 (0.99) 0.91 (1.00) 5. I felt hopeful about the future** 1.53 (1.09) 1.55 (1.10) 6. I felt fearful 0.73 (0.93) 0.55 (0.86) 7. My sleep was restless 0.84 (1.00) 0.93 (1.03) 8. I was happy 0.91 (0.98) 0.95 (1.97) 9. I felt lonely 0.75 (0.99) 1.04 (1.09) 10. I could not get “going” 0.52 (0.84) 0.66 (0.95) Before examining the longitudinal invariance, we investigated the structure of the scale in paper and online surveys separately. The results showed a model with poor fit for both modes of administration (CFI = 0.895, RMSEA = 0.125, SRMR = 0.062 for paper, and CFI = 0.866, RMSEA = 0.118, SRMR = 0.067 for the online survey). We explored modifications indices and standardized residuals in the Mplus output. The model was enhanced by permitting correlations among residuals of item 5 (I felt hopeful about the future) and 8 (I was happy) within time points. Table 2 indicates standardized item-by-item factor loadings observed across the two years of the study. The findings reveal a consistent and comparable pattern of factor loadings for the scale items, indicating a high degree of similarity in the factor structures over the two survey administrations. The new models with these correlations showed a good fit for both administration modes (Table 3 ). Table 2 Standardized factor loading of CESD-R-10 scale items in paper and online administration modes among COMPASS study (2018-19 and 2019-20) participants. Item Paper (2018-19) Online (2019-20) 1. I was bothered by things that usually don’t bother me 0.585 (0.018) 0.575 (0.018) 2. I had trouble keeping my mind on what I was doing 0.613 (0.016) 0.604 (0.016) 3. I felt depressed 0.798 (0.011) 0.743 (0.012) 4. I felt that everything I did was an effort 0.530 (0.019) 0.658 (0.015) 5. I felt hopeful about the future 0.226 (0.021) 0.168 (0.021) 6. I felt fearful 0.644 (0.016) 0.629 (0.016) 7. My sleep was restless 0.539 (0.018) 0.551 (0.017) 8. I was happy 0.453 (0.021) 0.358 (0.022) 9. I felt lonely 0.731 (0.013) 0.695 (0.013) 10. I could not get “going” 0.632 (0.017) 0.638 (0.016) Table 3 Measurement invariance of the CESD-R-10 scale across administration mode in two years (2018-19 and 2019-20) among COMPASS study participants Mode (School year) χ 2 (df) CFI SRMR RMSEA (90% CI) -∆CFI ∆SRMR ∆RMSEA Paper (2018-19) 414.2 (35) 0.975 0.034 0.062 (0.057, 0.067) -- -- -- Online (2019-20) 411.7 (35) 0.974 0.033 0.062 (0.056, 0.067) -- -- -- Mode analysis Configural 989.5 (157) 0.973 0.034 0.043 (0.041, 0.046) -- -- -- Metric 1002.9 (156) 0.972 0.033 0.044 (0.041, 0.046) 0.001 0.001 0.001 Scalar 1345.6 (186) 0.963 0.035 0.047 (0.045, 0.049) 0.009 0.002 0.003 Strict 1426.3 (196) 0.961 0.039 0.046 (0.045, 0.048) 0.002 0.004 0.001 Once this basic model was determined, we tested the configural invariance. An initial configural invariance model was specified, wherein a single-factor model was concurrently estimated in each administration mode. Model identification was achieved by fixing the factor variance to 1, the factor mean to zero, and constraining all residual variances to 1. Table 3 shows that the model had a good fit to the data, indicating the same structure of the scale across paper and online surveys. Equality of the item factor loadings between the two data collection modes was then examined in a metric invariance model. The factor variance was fixed to 1 in the paper mode for identification but was freely estimated in the online survey. After restricting the factor loadings to equality in both modes, the fit indices still indicated a good fit to the data. The restricted model showed a nonsignificant decrease in fit indices compared to the configural model ( ∆ CFI = 0.001, ∆SRMA = 0.001, ∆ RMSEA = 0.001). We tested the scalar invariance by restricting item intercepts and factor loadings to be invariant across the two time points. The fit criteria of the scalar model for the measures were met, showing support for the scalar variance. In the last stage, to evaluate the strict invariance, we constrained error variances to be equal over the two years. The results indicate the differences in fit indices met the criteria, indicating support for strict invariance. Having confirmed complete measurement invariance within our dataset, we proceeded to assess invariance across different gender identities among the adolescent sample (Table 4 ). The results indicate full measurement invariance for male students and partial measurement invariance for females. In males, the configural models, without any equality constraints, fit well to the data (CFI = 0.981; RMSEA = 0.033 [90% CI, 0.028–0.038], SRMR = 0.036). When constraints were applied at the metric, scalar, and strict stages, they did not substantially deteriorate the model’s fit (Table 3 ). For females, however, changes in criteria were slightly larger when imposing restrictions, especially the change from a metric to scalar model. We explored the modification indices to find the potential items that may have contributed to the high values of changes in measures. After relaxing the equality of intercepts on items 6 (I felt fearful) and item 9 (I felt lonely), the ∆CFI reduced to the acceptable level 0.007, indicating that it now meets the criteria. The strict invariance for females was also supported by the minimal changes in the measures in the model with constraints on all residuals and thresholds, except for item 6 and item 9 thresholds. Table 4 Measurement invariance of the CESD-R-10 scale across administration mode in two years (2018-19 and 2019-20) among COMPASS study participants Gender Mode (school year) χ 2 (df) CFI SRMR RMSEA (90% CI) -∆CFI ∆SRMR ∆RMSEA Female Paper (2018-19) 277.6 (34) 0.943 0.034 0.063 (0.056, 0.070) -- -- -- Online (2019-20) 261.9 (34) 0.961 0.034 0.061 (0.054, 0.068) -- -- -- Mode analysis Configural 875.5 (157) 0.963 0.040 0.050 (0.047, 0.054) -- -- -- Metric 832.5 (156) 0.965 0.037 0.049 (0.046, 0.052) 0.002 0.003 0.001 Scalar (equal intercept constrain for all items) 1225.9 (186) 0.946 0.040 0.056 (0.053, 0.059) 0.019 0.003 0.007 Scalar (equal intercept constrain removed for item 6 and item 9) 990.6 (180) 0.958 0.039 0.050 (0.047, 0.053) 0.007 0.002 0.001 Strict (equal intercept constrain removed for item 6 and item 9) 1026.8 (190) 0.957 0.044 0.049 (0.046, 0.052) 0.001 0.005 0.001 Male Paper (2018-19) 99.2 (34) 0.969 0.034 0.043 (0.034, 0.054) -- -- -- Online (2019-20) 116.7 (34) 0.957 0.034 0.049 (0.039, 0.059) -- -- -- Mode analysis Configural 332.5 (157) 0.981 0.036 0.033 (0.028, 0.038) -- -- -- Metric 341.8 (156) 0.980 0.036 0.034 (0.029, 0.039) 0.001 0.000 0.001 Scalar (equal intercept constrain for all items) 376.6 (186) 0.980 0.038 0.032 (0.027, 0.036) 0.000 0.002 0.002 Strict (equal intercept constrain for all items) 436.6 (196) 0.974 0.042 0.035 (0.030, 0.039) 0.006 0.004 0.003 CESD-R-10 scores in two administrations The average CESD-R-10 score in the paper survey (year 2019/20) was 8.19, lower than the online survey (year 2020/21) score of 8.98, indicating a score change of 0.79. The change in score varied among gender groups, with gender-diverse individuals showing a higher increase (2.18) compared to females (0.87) and males (0.64). Due to partial invariance of the scale among females, we investigated whether removing items 6 and 9 from the scale would affect our interpretation of depression score changes between the two years. Thus, we compared the scores from 2018 to 2019 using all 10 items versus excluding items 6 and 9 among females. The results revealed a 0.87 score change with the full scale and 0.78 with the partial scale. Overall, the interpretation of score changes from 2019/20 to 2020/21 using either score showed a significant increase (p < 0.001). The magnitude of change represented by the full score (0.87) was not different from that shown by the partial score (0.78, p = 0.051). Discussion We investigated the measurement invariance of a commonly used 10-item scale for depression symptoms over two administrations of a survey when data collection mode changed from a paper-based to an online administrated survey due to the COVID-19 pandemic onset among a large cohort of secondary school students. Consistent with past research (Motl et al., 2005 ; Verhoeven et al., 2013 ), our findings provide robust evidence supporting the consistent measurement properties of the CESD-R-10 among Canadian adolescents, reinforcing its reliability for assessing depressive symptoms over time. This confirmation is crucial for facilitating accurate and meaningful comparisons in longitudinal study designs when there has been a transition from a paper-based survey to an online format. In exploring the measurement invariance of the depression scale across administration modes, our initial model encountered challenges in accepting scalar invariance, suggesting potential differences in the intercepts across time points. This poor fit may partly be attributed to "pencil-whipping," where respondents select the same answer for all items without reading them. A high proportion of such responses can create the appearance of invariance (Meade & Craig, 2012 ). To address this concern, we investigated and identified thresholds of two items for which the equal threshold over the two administrations was not established in our dataset. Consistent with past research (Bagheri et al., 2021 ), we found the largest modification indices were associated with “I felt fearful” and “I felt lonely”. The decision to relax the equality constraint for these items proved pivotal. This adjustment remarkably improved the model fit and successfully established scalar invariance, implying that the majority of item thresholds were invariant across the administration assessments. The variance in the thresholds of these items may, to some extent, be attributed to the impact of the pandemic on participants' perceptions of the emotion of fear and loneliness. That is, the meaning of fear may have evolved under varying circumstances and contexts. The fears experienced during the pandemic, such as health concerns, academic disruptions, isolation, loneliness, and future uncertainty, may constitute genuine fears that did not exist before the pandemic. Considering the physical distancing measures and reduced social interactions among adolescents during the pandemic, variations in the interpretation and actual experience of loneliness between the two scale measurements can be anticipated. This nuanced finding highlights the sensitivity of measurement invariance tests to specific items and underscores the importance of scrutinizing individual intercepts in achieving a comprehensive understanding of the stability of the scale over time. The successful achievement of scalar invariance ensures that differences in observed scores between two administrations of CESD-R-10 accurately reflect genuine variations in the latent construct. Notably, our research revealed a gender-related variation in measurement invariance, particularly in scalar measurement invariance, which appeared to show full invariance among male adolescents while showing partially invariance among females, as reported in other studies (Verhoeven et al., 2013 ). This observation underscores the importance of recognizing potential gender-related differences in measurement properties, as these variations may influence the interpretation of results and have significant implications for research and applications. While some research suggests that a single unequal intercept can significantly impact the composite score, necessitating full scalar variance (Steinmetz, 2013 ), others suggest that partial scalar invariance may be adequate for intercept mean comparisons if the deviation from invariance is minor (Schmitt & Kuljanin, 2008 ). Considering the instability of the item "I felt fearful" and “I felt lonely” among females, consistent with recommendations in the study by Millsap and Kwok ( 2004 ), we suggest conducting a sensitivity analysis to compare the scale means from the administrations modes, both with and without these items. If no significant difference is found, researchers can utilize the total 10 items in their analysis. However, if a significant difference is observed, it may be advisable to report the results from the scale without including these specific items. Our sensitivity analysis revealed no significant difference between the full and partial scores, indicating use of full scale for depression score. However, researchers may benefit from conducting similar sensitivity analysis in their data to determine whether to use full or partial scores. The established measurement invariance of the CESD-R-10 when administered via a paper survey or an online survey holds importance for public health and research. Many decisions rely on outcomes derived from such tools. Public health decision makers require studies that employ measures that have demonstrated validity and reliability in the population of interest and the data collection mode implemented. This necessitates a rigorous process to determine structural validity, including testing the structural equivalence of data obtained from different measurements, allowing for inferences about depression trajectories and intervention effects. Our findings support the validity of findings using the CESD-R-10 to assess depression symptoms among adolescents and in prospective studies that changed data collection modes, as became increasingly relevant with advances in technology, the advantages of online surveys, and the pandemic onset. Results should be interpreted with consideration of the study limitations. First, a major limitation of the study is the inability to separate the effects of switching to an online survey mode from COVID-19 influences, as participants did not complete both paper and online versions at the same time. Second, COMPASS was not designed to be nationally representative, and thus, results may not be generalizable to all adolescents in Canada. However, the large sample size and use of active-information, passive-consent data collection protocols help to enhance the reliability and validity of our results. Third, due to a limited sample size, we were unable to analyze measurement invariance for gender diverse adolescents, a high priority given that this group had the highest depression scores. Also, adolescents that indicated they “prefer not to say” for were classified in the gender diverse category given the small sample size. The gender measure may conflate sex and gender. Future research using the revised sex and gender measures introduced in COMPASS will allow analyses across and within cisgender and gender diverse populations. LAstly, only participants with linked data were considered in this study, which means that grade 12 students and other students who were not linked over the course of the two-year study were not included in our analysis. In conclusion, our study provides evidence to support the measurement invariance of the CESD-R-10 in the context of changes in data collection methods. Findings demonstrate that the CESD-R-10 remains stable even with a shift in data collection mode from paper to online surveys. This provides support for the validity of assessing trajectories and conducting meaningful comparisons of depression symptoms over time in adolescent studies using longitudinal designs. Researchers considering changes in data collection mode can be confident in the consistency of the tool. Declarations Funding: The COMPASS study has been supported by a bridge grant from the CIHR Institute of Nutrition, Metabolism and Diabetes (INMD) through the “Obesity – Interventions to Prevent or Treat” priority funding awards (OOP-110788; awarded to SL), an operating grant from the CIHR Institute of Population and Public Health (IPPH) (MOP-114875; awarded to SL), a CIHR project grant (PJT-148562; awarded to SL), a CIHR bridge grant (PJT-149092; awarded to KP/SL), a CIHR project grant (PJT-159693; awarded to KP), and by a research funding arrangement with Health Canada (#1617-HQ-000012; contract awarded to SL) , a CIHR-Canadian Centre on Substance Abuse (CCSA) team grant (OF7 B1-PCPEGT 410-10-9633; awarded to SL), and a SickKids Foundation New Investigator Grant, in partnership with CIHR Institute of Human Development, Child and Youth Health (IHDCYH) (Grant No. NI21-1193; awarded to KAP) funds a mixed methods study examining the impact of the COVID-19 pandemic on youth mental health, leveraging COMPASS study data. The COMPASS-Quebec project additionally benefits from funding from the Ministère de la Santé et des Services sociaux of the province of Québec, and the Direction régionale de santé publique du CIUSSS de la Capitale-Nationale. MAF and KAP are supported the Canada Research Chairs program. Conflict of Interest: JM is a principal in BEAM Diagnostics, Inc. and a Consultant to Clairvoyant Therapeutics, Inc. The authors have no other conflicts of interest relevant to this article to disclose. Research Data Statement: The COMPASS study data can be accesses by obtaining approval through an online form found at: https://uwaterloo.ca/compass-system/information-researchers/data-usage-application References Andresen EM, Malmgren JA, Carter WB, Patrick DL (1994) Screening for depression in well older adults: Evaluation of a short form of the CES-D. Am J Prev Med 10(2):77–84 Bagheri Z, Noorshargh P, Shahsavar Z, Jafari P (2021) Assessing the measurement invariance of the 10-item Centre for Epidemiological Studies Depression Scale and Beck Anxiety Inventory questionnaires across people living with HIV/AIDS and healthy people. BMC Psychol 9:1–11 Becker R (2022) Gender and survey participation: an event history analysis of the gender effects of survey participation in a probability-based multi-wave panel study with a sequential mixed-mode design. methods, data, analyses , 16 (1), 30 Bitsko R, Claussen A, Lichstein J, Black L, Everett Jones S, Danielson M, Hoenig J, Jack SD, Brody D, Gyawali S (2022) Surveillance of children’s mental health–United States, 2013–2019. MMWR Supplements 71(2):1–42 Chen FF (2007) Sensitivity of goodness of fit indexes to lack of measurement invariance. Struct Equation Modeling: Multidisciplinary J 14(3):464–504 Clayborne ZM, Varin M, Colman I (2019) Systematic review and meta-analysis: adolescent depression and long-term psychosocial outcomes. J Am Acad Child Adolesc Psychiatry 58(1):72–79 Glied S, Pine DS (2002) Consequences and correlates of adolescent depression. Arch Pediatr Adolesc Med 156(10):1009–1014 Haroz EE, Ybarra ML, Eaton WW (2014) Psychometric evaluation of a self-report scale to measure adolescent depression: the CESDR-10 in two national adolescent samples in the United States. J Affect Disord 158:154–160 Jonsson U, Bohman H, von Knorring L, Olsson G, Paaren A, Von Knorring A-L (2011) Mental health outcome of long-term and episodic adolescent depression: 15-year follow-up of a community sample. J Affect Disord 130(3):395–404 Kline RB (2023) Principles and practice of structural equation modeling. Guilford Leatherdale ST, Brown KS, Carson V, Childs RA, Dubin JA, Elliott SJ, Faulkner G, Hammond D, Manske S, Sabiston CM (2014) The COMPASS study: a longitudinal hierarchical research platform for evaluating natural experiments related to changes in school-level programs, policies and built environment resources. BMC Public Health 14(1):1–7 Meade AW, Craig SB (2012) Identifying careless responses in survey data. Psychol Methods 17(3):437 Millsap RE, Kwok O-M (2004) Evaluating the impact of partial factorial invariance on selection in two populations. Psychol Methods 9(1):93 Mossman SA, Luft MJ, Schroeder HK, Varney ST, Fleck DE, Barzman DH, Gilman R, DelBello MP, Strawn JR (2017) The Generalized Anxiety Disorder 7-item (GAD-7) scale in adolescents with generalized anxiety disorder: signal detection and validation. Annals Clin psychiatry: official J Am Acad Clin Psychiatrists 29(4):227 Motl RW, Dishman RK, Birnbaum AS, Lytle LA (2005) Longitudinal invariance of the Center for Epidemiologic Studies-Depression Scale among girls and boys in middle school. Educ Psychol Meas 65(1):90–108 Muthen LK, Muthen B, Muthén,M (2017) Mplus Version 8 User's Guide. Muthen & Muthen. https://books.google.ca/books?id=dgDlAQAACAAJ Radloff LS (1977) The CES-D scale: A self-report depression scale for research in the general population. Appl Psychol Meas 1(3):385–401 Romano I, Ferro MA, Patte KA, Leatherdale ST (2021) Measurement invariance of the GAD-7 and CESD-R-10 among adolescents in Canada. J Pediatr Psychol Schmitt N, Kuljanin G (2008) Measurement invariance: Review of practice and implications. Hum resource Manage Rev 18(4):210–222 Steinmetz H (2013) Analyzing observed composite differences across groups. Methodology Verhoeven M, Sawyer MG, Spence SH (2013) The factorial invariance of the CES-D during adolescence: are symptom profiles for depression stable across gender and time? J Adolesc 36(1):181–190 Widaman KF, Reise SP (1997) Exploring the measurement invariance of psychological instruments: Applications in the substance use domain Zhang X, Kuchinke L, Woud ML, Velten J, Margraf J (2017) Survey method matters: Online/offline questionnaires and face-to-face or telephone interviews differ. Comput Hum Behav 71:172–180 Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5815737","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":401163040,"identity":"b366570a-5991-4c1a-8c1b-f899bd704d55","order_by":0,"name":"Mahmood R Gohari","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4ElEQVRIiWNgGAWjYHACNiCWkINyLIjXYgwkGBuADKK1MCQ2EK1Ft3/xsQc/91ikb7iR/PzBxz0ScvwMzA8/4NNiduNZumHPM4ncDTfSDBtnPJMwlmxgM8Zrl9mNM2YSPAeAWm4nGDYDGYkbDvDgdx5Ii+SfAxLpBrfTPzb/gWhh/oFXy/keM2mg4QkGt3MMmxkgWtgI2MKWJi1zQMJw5v03hTN7DgD90sxmhjd2zM4fPib55kCdPN+Z4xs+/DhgI8fP3vz4Bj4tDBIJ6CLMeNUDAf8BQipGwSgYBaNgxAMAtmBNQYOD2+kAAAAASUVORK5CYII=","orcid":"","institution":"University of Waterloo","correspondingAuthor":true,"prefix":"","firstName":"Mahmood","middleName":"R","lastName":"Gohari","suffix":""},{"id":401163041,"identity":"61ca1aa5-fa48-4f2e-b1bc-560df62ea2c6","order_by":1,"name":"Mark A. Ferro","email":"","orcid":"","institution":"University of Waterloo","correspondingAuthor":false,"prefix":"","firstName":"Mark","middleName":"A.","lastName":"Ferro","suffix":""},{"id":401163042,"identity":"2ed700b3-bb47-4b64-8f16-e913335863e6","order_by":2,"name":"Karen Patte","email":"","orcid":"","institution":"Brock University","correspondingAuthor":false,"prefix":"","firstName":"Karen","middleName":"","lastName":"Patte","suffix":""},{"id":401163043,"identity":"1393eb30-6c47-45e2-a774-99cfe28c0535","order_by":3,"name":"James MacKillop","email":"","orcid":"","institution":"McMaster University","correspondingAuthor":false,"prefix":"","firstName":"James","middleName":"","lastName":"MacKillop","suffix":""},{"id":401163044,"identity":"82df1d02-b08f-4215-a356-03c2e30520a5","order_by":4,"name":"Scott T Leatherdale","email":"","orcid":"","institution":"University of Waterloo","correspondingAuthor":false,"prefix":"","firstName":"Scott","middleName":"T","lastName":"Leatherdale","suffix":""}],"badges":[],"createdAt":"2025-01-13 00:48:30","currentVersionCode":1,"declarations":{"humanSubjects":true,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":true,"humanSubjectConsent":true,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-5815737/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5815737/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":73738038,"identity":"05c859c0-9168-4b10-9b1c-ce9a55a29ae0","added_by":"auto","created_at":"2025-01-14 07:36:04","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":645436,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5815737/v1/776e7995-51d3-4bdd-9263-967c815d0d47.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eMeasurement Invariance of the CESD-R-10 Among Adolescents over the Transition from Paper to Online Administration\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eAdolescent depression is a significant public health concern. Recent research indicated that more than 20% of adolescents aged 12\u0026ndash;17 years had experienced a major depressive episode and 36.7% reported consistent feelings of sadness or hopelessness within the past year (Bitsko et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The significance of adolescent depression arises from its potential to result in immediate and long-term consequences. Such consequences include disruptions in educational attainment, impaired social relationships, and increased risk of suicide and substance use (Clayborne et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Glied \u0026amp; Pine, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). The impact can extend into adulthood; those who experience depression during adolescence often face long-term psychosocial ramifications and reduced physical health and functioning in later life (Jonsson et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eResearch is essential to advance our understanding and inform more effective interventions that can alleviate the burden of depression in adolescents. To accomplish these aims, it is imperative to employ robust measures for assessing adolescent depression. One of the most widely used depression scales is the Center for Epidemiologic Studies Depression Scale (CESD) (Radloff, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1977\u003c/span\u003e) and its shorter form versions (e.g., the CESD-R-10) (Andresen et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1994\u003c/span\u003e). These scales have been used in surveys with different modes of administration. In particular, during the pandemic many surveys had to move from traditional paper-based methods to an online mode due to physical distancing measures. This shift may influence participant responses. For instance, anonymity and privacy concerns may differ between paper and online modes, potentially influencing comfort of providing honest responses. Moreover, interactive elements in online surveys may influence how participants engage with and respond to questions compared to static paper-based questionnaires.\u003c/p\u003e \u003cp\u003eAside from the impact of the COVID-19 pandemic, online data collection has become a commonly used method due to its lower cost, time efficiency, enhanced accessibility, and increased flexibility, all enabled by advancements in technology. The ease of administration, diverse question types, multimedia integration, and adaptive questioning based on specific responses further contribute to the appeal of online surveys. Moreover, online surveys align with environmentally friendly practices by reducing reliance on paper, printing, and transportation. Considering adolescents\u0026rsquo; ubiquitous use of and comfort with screen- and web-based platforms, surveys targeting this demographic could be more inclined to opt for online modes of data collection.\u003c/p\u003e \u003cp\u003eDespite potential differences in performance between paper and online surveys, there remains a notable dearth of research validating the measurement properties of scales originally designed for paper administration when applied to online methodologies. To ensure reliable and valid conclusions, it is essential to demonstrate that measurement tools maintain the same factorial structure and assess the same construct across different administration modes. Measurement invariance is important, in particular, for longitudinal surveys that aim to examine trajectories of depression symptoms over time and have a change in data collection mode. Ensuring appropriate comparisons of a construct before and after administration mode change relies on establishing the equivalence and consistency of its meaning. Without this consistency, drawing comparisons and exploring trends may lead to invalid conclusions.\u003c/p\u003e \u003cp\u003eGiven the importance of measurement consistency when changing mode of data collection, the main objective of the current study is to evaluate the invariance of a 10-item version of the CESD across paper-based and online surveys conducted pre- and post- COVID-19 pandemic onset in a large sample of adolescents. Previous research has demonstrated that the survey method can notably impact the responses of gender groups to certain psychometric scales (e.g., happiness) (Zhang et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Therefore, considering the observed gender differences in depression symptoms and their engagement in online surveys (Becker, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), a secondary purpose of this paper is to consider whether associations differ between gender groups.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eDesign and participation\u003c/h2\u003e \u003cp\u003eThe longitudinal measurement invariance of the CESD-R-10 was studied across two repeated measurements among adolescents who participated in the COMPASS (Cannabis, Obesity, Mental health, Physical activity, Alcohol, Smoking, and Sedentary behaviour) study (Leatherdale et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). The sample included 2,831 secondary school students who provided two-year linked data in the 2019/20 (T1) and 2020/21 (T2) school years. Student data in T1 were collected using a paper-based machine-readable survey completed during one classroom period (Reel et al., 2020). As the COVID-19 pandemic was declared in Canada in March 2020 and schools were closed to in-person learning, T2 data were collected online using the Qualtrics XM survey software (Qualtrics, Provo, UT, USA). Student data were linked across the study waves utilizing an anonymous unique identification code, which was generated based on information derived from five initial questionnaire questions. Full school samples of students were invited to participate using an active-information passive-consent permission protocol.\u003c/p\u003e \u003cp\u003e All the processes were granted ethical approval by the University of Waterloo (ORE#30118), Brock University (REB#18\u0026ndash;099), CIUSSS de la Capitale-Nationale\u0026ndash;Universit\u0026eacute; Laval (#MP-13-2017-1264), and the school boards involved.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eMeasures\u003c/h3\u003e\n\u003cp\u003eDepression was assessed using a 10-item version of the CESD (CESD-R-10) (Andresen et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1994\u003c/span\u003e). The CESD-R-10 uses a 4-point response format (0 = \"None or less than 1 day\"; 1 = \"1\u0026ndash;2 days\"; 2 = \"3\u0026ndash;4 days\"; 3 = \"5\u0026ndash;7 days\") for participants to indicate the frequency to which they experienced different symptoms of unipolar depression in the last 7 days. Scores across the items were summed to create a total score ranging from a possible 0 to 30; higher scores indicated more and more frequent depression symptoms and a higher risk of clinically-relevant major depressive symptomatology. In this study, the internal consistency across CESD-R-10 items was 0.83 and 0.82 in the paper and online administered surveys, respectively. The validity and reliability of CESD-R-10 in adolescent populations have been documented (Haroz et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Mossman et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), including cross-sectional measurement invariance by sex and grade in the COMPASS study using paper administration (Romano et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eGender: Students were asked \u0026ldquo;are you a female or male?\u0026rdquo;, with response options including female, male, I describe my gender in a different way, and I prefer not to say. Due to the limited sample size of the last two options, responses were grouped into female, male, and gender-diverse/prefer not to say categories.\u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eLongitudinal measurement invariance analyses were conducted within a structural equation modeling framework over the study period. Different levels of invariance were evaluated through successive comparisons of measurement models, progression from the least to the most restrictive model as outlined below (Widaman \u0026amp; Reise, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e1997\u003c/span\u003e):\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eConfigural Invariance: This stage examined whether the general structure of the measurement model was the same across administrations. Therefore, a paper administration CFA model (2018-19 school year) was compared with models from the online survey (2019-20) to investigate configural invariance.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eMetric Invariance: This stage examined the equality of factor loadings across survey administration mode. The metric invariance aimed to ascertain if the scale items possessed the same meaning and strength of connection with the underlying depression construct during both measurements.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eScalar Invariance: Scalar invariance was evaluated by imposing equal constraints on both factor loadings and item thresholds across the two time points. Scalar invariance implies that the scale maintains consistent measurement units and thresholds over time, and the relationship between the latent variable and its indicators remains unchanged despite change in the mode of survey administration.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eStrict Invariance: Further constrained added to the previous model by requiring equal residual variances across paper and online surveys to test the strict invariance.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eMeasurement invariance was attained by assessing all four stages. If the model fit statistics were favorable at the initial stage of configural invariance, subsequent levels of the model were constrained. If the constrained values yielded a fit that was not inferior to the initial model, it was decided that the stage of measurement invariance has been attained. The Weighted Least Squares Mean and Variance adjusted (WLSMV) including THETA parametrization was used in estimating factor models because the response options were on Likert scale (Muthen et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWe employed two criteria to ensure measurement equivalence. The initial criterion involved ensuring that the model fit was satisfactory at every testing level. To assess the adequacy of model fit, we used the comparative fit index (CFI) with a threshold of \u0026ge;\u0026thinsp;0.90 and the root mean square error of approximation (RMSEA) and standardized root mean square residuals (SRMR) with a threshold of \u0026le;\u0026thinsp;0.06 (Kline, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The adequacy of the fit is considered satisfactory if the criteria are met by at least two of these fit indices. No statistically significant deterioration was anticipated when constraints were imposed to a model, assuming that invariance was accomplished. As such, the second criterion specified that any changes in fit indices (i.e., when moving from a model with fewer equality constraints on parameters to a more constrained model) should not exceed pre-established thresholds. We predetermined that, to establish measurement equivalence at a particular testing stage, at least two of the following fit indices, namely difference in fit indices between the two time points as ΔCFI, ΔRMSEA, or ΔSRMR, needed to meet this criterion. The defined cutoff values for changes in model fit indices were set at ΔCFI\u0026thinsp;\u0026le;\u0026thinsp;0.010, ΔRMSEA\u0026thinsp;\u0026le;\u0026thinsp;0.015, or ΔSRMR\u0026thinsp;\u0026le;\u0026thinsp;0.030 (Chen, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2007\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn cases where measurement invariance at a specific testing level was not achieved (signified by a notable deterioration in model fit), we examined partial measurement invariance. This examination aimed to identify constraints on relevant non-invariant parameters that could be removed to enhance the overall model fit. Subsequently, we tested partial invariance by comparing fit of the originally proposed model against the less constrained version. In addition, modifications indices and standardized residuals in the Mplus output were examined to enhance the model fit. We examined measurement invariance by gender groups. However, due to the small sample of students who self-identified as gender diverse, the stratified analysis was only conducted for males and females. Data management was performed with SAS Studio [Enterprise Edition]. Mplus version 8.4 was used for the confirmatory factor analyses and measurement invariance analyses.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eOver half of students (63.2%) identified as females and 1% were classified as gender diverse. At baseline (2018-19), students\u0026rsquo; mean age was 14.3 years (SD:1.2, range: 12 to 17 years) and 62.0% were within grades 9\u0026ndash;11. The mean depression score at baseline was 8.2 (SD\u0026thinsp;=\u0026thinsp;5.9) and 33.7% (N\u0026thinsp;=\u0026thinsp;953) of students reported clinically-relevant depression symptoms (CESD-R-10 score\u0026thinsp;\u0026ge;\u0026thinsp;10). Gender diverse students reported a higher mean depression score (11.9) than their female (9.1) and male (6.5) peers. Mean scores of scale items in the two years are presented in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCESD-R-10 Scale item means (SD) by administration mode in two years among COMPASS study (2018-19 and 2019-20) participants. *Item responses were coded as 0 (none or less than 1 day) to 3 (5\u0026ndash;7 days). **Items 5 and 8 were reverse coded to be consistent with the others. Higher scores indicate more depression symptoms.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eItem\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePaper (2018-19)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOnline (2019-20)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1. I was bothered by things that usually don\u0026rsquo;t bother me*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.43 (0.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.50 (0.79)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2. I had trouble keeping my mind on what I was doing\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.87 (0.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.04 (1.02)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3. I felt depressed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.74 (0.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.84 (1.00)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4. I felt that everything I did was an effort\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.85 (0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.91 (1.00)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5. I felt hopeful about the future**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.53 (1.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.55 (1.10)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6. I felt fearful\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.73 (0.93)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.55 (0.86)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7. My sleep was restless\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.84 (1.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.93 (1.03)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8. I was happy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.91 (0.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.95 (1.97)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9. I felt lonely\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.75 (0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.04 (1.09)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10. I could not get \u0026ldquo;going\u0026rdquo;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.52 (0.84)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.66 (0.95)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eBefore examining the longitudinal invariance, we investigated the structure of the scale in paper and online surveys separately. The results showed a model with poor fit for both modes of administration (CFI\u0026thinsp;=\u0026thinsp;0.895, RMSEA\u0026thinsp;=\u0026thinsp;0.125, SRMR\u0026thinsp;=\u0026thinsp;0.062 for paper, and CFI\u0026thinsp;=\u0026thinsp;0.866, RMSEA\u0026thinsp;=\u0026thinsp;0.118, SRMR\u0026thinsp;=\u0026thinsp;0.067 for the online survey). We explored modifications indices and standardized residuals in the Mplus output. The model was enhanced by permitting correlations among residuals of item 5 (I felt hopeful about the future) and 8 (I was happy) within time points. Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e indicates standardized item-by-item factor loadings observed across the two years of the study. The findings reveal a consistent and comparable pattern of factor loadings for the scale items, indicating a high degree of similarity in the factor structures over the two survey administrations. The new models with these correlations showed a good fit for both administration modes (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eStandardized factor loading of CESD-R-10 scale items in paper and online administration modes among COMPASS study (2018-19 and 2019-20) participants.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eItem\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePaper (2018-19)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOnline (2019-20)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1. I was bothered by things that usually don\u0026rsquo;t bother me\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.585 (0.018)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.575 (0.018)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2. I had trouble keeping my mind on what I was doing\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.613 (0.016)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.604 (0.016)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3. I felt depressed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.798 (0.011)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.743 (0.012)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4. I felt that everything I did was an effort\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.530 (0.019)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.658 (0.015)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5. I felt hopeful about the future\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.226 (0.021)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.168 (0.021)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6. I felt fearful\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.644 (0.016)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.629 (0.016)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7. My sleep was restless\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.539 (0.018)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.551 (0.017)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8. I was happy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.453 (0.021)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.358 (0.022)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9. I felt lonely\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.731 (0.013)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.695 (0.013)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10. I could not get \u0026ldquo;going\u0026rdquo;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.632 (0.017)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.638 (0.016)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMeasurement invariance of the CESD-R-10 scale across administration mode in two years (2018-19 and 2019-20) among COMPASS study participants\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMode (School year)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eχ\u003csup\u003e2\u003c/sup\u003e (df)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCFI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSRMR\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRMSEA (90% CI)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-∆CFI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e∆SRMR\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e∆RMSEA\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePaper (2018-19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e414.2 (35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.975\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.034\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.062 (0.057, 0.067)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOnline (2019-20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e411.7 (35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.974\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.033\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.062 (0.056, 0.067)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMode analysis\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConfigural\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e989.5 (157)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.973\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.034\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.043 (0.041, 0.046)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetric\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1002.9 (156)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.972\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.033\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.044 (0.041, 0.046)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScalar\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1345.6 (186)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.963\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.035\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.047 (0.045, 0.049)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.009\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStrict\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1426.3 (196)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.961\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.039\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.046 (0.045, 0.048)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eOnce this basic model was determined, we tested the configural invariance. An initial configural invariance model was specified, wherein a single-factor model was concurrently estimated in each administration mode. Model identification was achieved by fixing the factor variance to 1, the factor mean to zero, and constraining all residual variances to 1. Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e shows that the model had a good fit to the data, indicating the same structure of the scale across paper and online surveys.\u003c/p\u003e \u003cp\u003eEquality of the item factor loadings between the two data collection modes was then examined in a metric invariance model. The factor variance was fixed to 1 in the paper mode for identification but was freely estimated in the online survey. After restricting the factor loadings to equality in both modes, the fit indices still indicated a good fit to the data. The restricted model showed a nonsignificant decrease in fit indices compared to the configural model (\u003cb\u003e∆\u003c/b\u003eCFI\u0026thinsp;=\u0026thinsp;0.001, ∆SRMA\u0026thinsp;=\u0026thinsp;0.001, \u003cb\u003e∆\u003c/b\u003eRMSEA\u0026thinsp;=\u0026thinsp;0.001). We tested the scalar invariance by restricting item intercepts and factor loadings to be invariant across the two time points. The fit criteria of the scalar model for the measures were met, showing support for the scalar variance. In the last stage, to evaluate the strict invariance, we constrained error variances to be equal over the two years. The results indicate the differences in fit indices met the criteria, indicating support for strict invariance.\u003c/p\u003e \u003cp\u003eHaving confirmed complete measurement invariance within our dataset, we proceeded to assess invariance across different gender identities among the adolescent sample (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The results indicate full measurement invariance for male students and partial measurement invariance for females. In males, the configural models, without any equality constraints, fit well to the data (CFI\u0026thinsp;=\u0026thinsp;0.981; RMSEA\u0026thinsp;=\u0026thinsp;0.033 [90% CI, 0.028\u0026ndash;0.038], SRMR\u0026thinsp;=\u0026thinsp;0.036). When constraints were applied at the metric, scalar, and strict stages, they did not substantially deteriorate the model\u0026rsquo;s fit (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). For females, however, changes in criteria were slightly larger when imposing restrictions, especially the change from a metric to scalar model. We explored the modification indices to find the potential items that may have contributed to the high values of changes in measures. After relaxing the equality of intercepts on items 6 (I felt fearful) and item 9 (I felt lonely), the ∆CFI reduced to the acceptable level 0.007, indicating that it now meets the criteria. The strict invariance for females was also supported by the minimal changes in the measures in the model with constraints on all residuals and thresholds, except for item 6 and item 9 thresholds.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMeasurement invariance of the CESD-R-10 scale across administration mode in two years (2018-19 and 2019-20) among COMPASS study participants\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMode (school year)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eχ\u003csup\u003e2\u003c/sup\u003e (df)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCFI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSRMR\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRMSEA (90% CI)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-∆CFI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e∆SRMR\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003e∆RMSEA\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"7\" rowspan=\"8\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePaper (2018-19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e277.6 (34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.943\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.034\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.063 (0.056, 0.070)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOnline (2019-20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e261.9 (34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.961\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.034\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.061 (0.054, 0.068)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"8\" nameend=\"c9\" namest=\"c2\"\u003e \u003cp\u003e\u003cb\u003eMode analysis\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eConfigural\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e875.5 (157)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.963\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.040\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.050 (0.047, 0.054)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMetric\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e832.5 (156)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.965\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.037\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.049 (0.046, 0.052)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eScalar (equal intercept constrain for all items)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1225.9 (186)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.946\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.040\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.056 (0.053, 0.059)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eScalar (equal intercept constrain removed for item 6 and item 9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e990.6 (180)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.958\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.039\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.050 (0.047, 0.053)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eStrict (equal intercept constrain removed for item 6 and item 9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1026.8 (190)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.957\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.044\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.049 (0.046, 0.052)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"6\" rowspan=\"7\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePaper (2018-19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e99.2 (34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.969\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.034\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.043 (0.034, 0.054)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOnline (2019-20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e116.7 (34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.957\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.034\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.049 (0.039, 0.059)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"8\" nameend=\"c9\" namest=\"c2\"\u003e \u003cp\u003e\u003cb\u003eMode analysis\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eConfigural\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e332.5 (157)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.981\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.033 (0.028, 0.038)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e--\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMetric\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e341.8 (156)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.980\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.034 (0.029, 0.039)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eScalar (equal intercept constrain for all items)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e376.6 (186)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.980\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.038\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.032 (0.027, 0.036)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eStrict (equal intercept constrain for all items)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e436.6 (196)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.974\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.042\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.035 (0.030, 0.039)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.006\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e\n\u003ch3\u003eCESD-R-10 scores in two administrations\u003c/h3\u003e\n\u003cp\u003eThe average CESD-R-10 score in the paper survey (year 2019/20) was 8.19, lower than the online survey (year 2020/21) score of 8.98, indicating a score change of 0.79. The change in score varied among gender groups, with gender-diverse individuals showing a higher increase (2.18) compared to females (0.87) and males (0.64). Due to partial invariance of the scale among females, we investigated whether removing items 6 and 9 from the scale would affect our interpretation of depression score changes between the two years. Thus, we compared the scores from 2018 to 2019 using all 10 items versus excluding items 6 and 9 among females. The results revealed a 0.87 score change with the full scale and 0.78 with the partial scale. Overall, the interpretation of score changes from 2019/20 to 2020/21 using either score showed a significant increase (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The magnitude of change represented by the full score (0.87) was not different from that shown by the partial score (0.78, p\u0026thinsp;=\u0026thinsp;0.051).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eWe investigated the measurement invariance of a commonly used 10-item scale for depression symptoms over two administrations of a survey when data collection mode changed from a paper-based to an online administrated survey due to the COVID-19 pandemic onset among a large cohort of secondary school students. Consistent with past research (Motl et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Verhoeven et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), our findings provide robust evidence supporting the consistent measurement properties of the CESD-R-10 among Canadian adolescents, reinforcing its reliability for assessing depressive symptoms over time. This confirmation is crucial for facilitating accurate and meaningful comparisons in longitudinal study designs when there has been a transition from a paper-based survey to an online format.\u003c/p\u003e \u003cp\u003eIn exploring the measurement invariance of the depression scale across administration modes, our initial model encountered challenges in accepting scalar invariance, suggesting potential differences in the intercepts across time points. This poor fit may partly be attributed to \"pencil-whipping,\" where respondents select the same answer for all items without reading them. A high proportion of such responses can create the appearance of invariance (Meade \u0026amp; Craig, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo address this concern, we investigated and identified thresholds of two items for which the equal threshold over the two administrations was not established in our dataset. Consistent with past research (Bagheri et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), we found the largest modification indices were associated with \u0026ldquo;I felt fearful\u0026rdquo; and \u0026ldquo;I felt lonely\u0026rdquo;. The decision to relax the equality constraint for these items proved pivotal. This adjustment remarkably improved the model fit and successfully established scalar invariance, implying that the majority of item thresholds were invariant across the administration assessments. The variance in the thresholds of these items may, to some extent, be attributed to the impact of the pandemic on participants' perceptions of the emotion of fear and loneliness. That is, the meaning of fear may have evolved under varying circumstances and contexts. The fears experienced during the pandemic, such as health concerns, academic disruptions, isolation, loneliness, and future uncertainty, may constitute genuine fears that did not exist before the pandemic. Considering the physical distancing measures and reduced social interactions among adolescents during the pandemic, variations in the interpretation and actual experience of loneliness between the two scale measurements can be anticipated. This nuanced finding highlights the sensitivity of measurement invariance tests to specific items and underscores the importance of scrutinizing individual intercepts in achieving a comprehensive understanding of the stability of the scale over time. The successful achievement of scalar invariance ensures that differences in observed scores between two administrations of CESD-R-10 accurately reflect genuine variations in the latent construct.\u003c/p\u003e \u003cp\u003eNotably, our research revealed a gender-related variation in measurement invariance, particularly in scalar measurement invariance, which appeared to show full invariance among male adolescents while showing partially invariance among females, as reported in other studies (Verhoeven et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). This observation underscores the importance of recognizing potential gender-related differences in measurement properties, as these variations may influence the interpretation of results and have significant implications for research and applications. While some research suggests that a single unequal intercept can significantly impact the composite score, necessitating full scalar variance (Steinmetz, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), others suggest that partial scalar invariance may be adequate for intercept mean comparisons if the deviation from invariance is minor (Schmitt \u0026amp; Kuljanin, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Considering the instability of the item \"I felt fearful\" and \u0026ldquo;I felt lonely\u0026rdquo; among females, consistent with recommendations in the study by Millsap and Kwok (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), we suggest conducting a sensitivity analysis to compare the scale means from the administrations modes, both with and without these items. If no significant difference is found, researchers can utilize the total 10 items in their analysis. However, if a significant difference is observed, it may be advisable to report the results from the scale without including these specific items. Our sensitivity analysis revealed no significant difference between the full and partial scores, indicating use of full scale for depression score. However, researchers may benefit from conducting similar sensitivity analysis in their data to determine whether to use full or partial scores.\u003c/p\u003e \u003cp\u003eThe established measurement invariance of the CESD-R-10 when administered via a paper survey or an online survey holds importance for public health and research. Many decisions rely on outcomes derived from such tools. Public health decision makers require studies that employ measures that have demonstrated validity and reliability in the population of interest and the data collection mode implemented. This necessitates a rigorous process to determine structural validity, including testing the structural equivalence of data obtained from different measurements, allowing for inferences about depression trajectories and intervention effects. Our findings support the validity of findings using the CESD-R-10 to assess depression symptoms among adolescents and in prospective studies that changed data collection modes, as became increasingly relevant with advances in technology, the advantages of online surveys, and the pandemic onset.\u003c/p\u003e \u003cp\u003eResults should be interpreted with consideration of the study limitations. First, a major limitation of the study is the inability to separate the effects of switching to an online survey mode from COVID-19 influences, as participants did not complete both paper and online versions at the same time. Second, COMPASS was not designed to be nationally representative, and thus, results may not be generalizable to all adolescents in Canada. However, the large sample size and use of active-information, passive-consent data collection protocols help to enhance the reliability and validity of our results. Third, due to a limited sample size, we were unable to analyze measurement invariance for gender diverse adolescents, a high priority given that this group had the highest depression scores. Also, adolescents that indicated they \u0026ldquo;prefer not to say\u0026rdquo; for were classified in the gender diverse category given the small sample size. The gender measure may conflate sex and gender. Future research using the revised sex and gender measures introduced in COMPASS will allow analyses across and within cisgender and gender diverse populations. LAstly, only participants with linked data were considered in this study, which means that grade 12 students and other students who were not linked over the course of the two-year study were not included in our analysis.\u003c/p\u003e \u003cp\u003eIn conclusion, our study provides evidence to support the measurement invariance of the CESD-R-10 in the context of changes in data collection methods. Findings demonstrate that the CESD-R-10 remains stable even with a shift in data collection mode from paper to online surveys. This provides support for the validity of assessing trajectories and conducting meaningful comparisons of depression symptoms over time in adolescent studies using longitudinal designs. Researchers considering changes in data collection mode can be confident in the consistency of the tool.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding:\u0026nbsp;\u003c/strong\u003eThe COMPASS study has been supported by a bridge grant from the CIHR Institute of Nutrition, Metabolism and Diabetes (INMD) through the \u0026ldquo;Obesity \u0026ndash; Interventions to Prevent or Treat\u0026rdquo; priority funding awards (OOP-110788; awarded to SL), an operating grant from the CIHR Institute of Population and Public Health (IPPH) (MOP-114875; awarded to SL), a CIHR project grant (PJT-148562; awarded to SL), a CIHR bridge grant (PJT-149092; awarded to KP/SL), a CIHR project grant (PJT-159693; awarded to KP), and by a research funding arrangement with Health Canada (#1617-HQ-000012; contract awarded to SL)\u003cem\u003e,\u003cem\u003e\u0026nbsp;a CIHR-Canadian Centre on Substance Abuse (CCSA) team grant (OF7 B1-PCPEGT 410-10-9633; awarded to SL), and a SickKids Foundation New Investigator Grant, in partnership with CIHR Institute of Human Development, Child and Youth Health (IHDCYH) (Grant No. NI21-1193; awarded to KAP) funds a mixed methods study examining the impact of the COVID-19 pandemic on youth mental health, leveraging COMPASS study data. The COMPASS-Quebec project additionally benefits from funding from the Minist\u0026egrave;re de la Sant\u0026eacute; et des Services sociaux of the province of Qu\u0026eacute;bec, and the Direction r\u0026eacute;gionale de sant\u0026eacute; publique du CIUSSS de la Capitale-Nationale. MAF and KAP are supported the Canada Research Chairs program.\u003c/em\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest:\u0026nbsp;\u003c/strong\u003eJM is a principal in BEAM Diagnostics, Inc. and a Consultant to Clairvoyant Therapeutics, Inc. The authors have no other conflicts of interest relevant to this article to disclose.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResearch Data Statement:\u003c/strong\u003e The COMPASS study data can be accesses by obtaining approval through an online form found at: https://uwaterloo.ca/compass-system/information-researchers/data-usage-application\u003c/p\u003e\n"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAndresen EM, Malmgren JA, Carter WB, Patrick DL (1994) Screening for depression in well older adults: Evaluation of a short form of the CES-D. Am J Prev Med 10(2):77\u0026ndash;84\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBagheri Z, Noorshargh P, Shahsavar Z, Jafari P (2021) Assessing the measurement invariance of the 10-item Centre for Epidemiological Studies Depression Scale and Beck Anxiety Inventory questionnaires across people living with HIV/AIDS and healthy people. BMC Psychol 9:1\u0026ndash;11\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBecker R (2022) Gender and survey participation: an event history analysis of the gender effects of survey participation in a probability-based multi-wave panel study with a sequential mixed-mode design. \u003cem\u003emethods, data, analyses\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(1), 30\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBitsko R, Claussen A, Lichstein J, Black L, Everett Jones S, Danielson M, Hoenig J, Jack SD, Brody D, Gyawali S (2022) Surveillance of children\u0026rsquo;s mental health\u0026ndash;United States, 2013\u0026ndash;2019. MMWR Supplements 71(2):1\u0026ndash;42\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen FF (2007) Sensitivity of goodness of fit indexes to lack of measurement invariance. Struct Equation Modeling: Multidisciplinary J 14(3):464\u0026ndash;504\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClayborne ZM, Varin M, Colman I (2019) Systematic review and meta-analysis: adolescent depression and long-term psychosocial outcomes. J Am Acad Child Adolesc Psychiatry 58(1):72\u0026ndash;79\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlied S, Pine DS (2002) Consequences and correlates of adolescent depression. Arch Pediatr Adolesc Med 156(10):1009\u0026ndash;1014\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaroz EE, Ybarra ML, Eaton WW (2014) Psychometric evaluation of a self-report scale to measure adolescent depression: the CESDR-10 in two national adolescent samples in the United States. J Affect Disord 158:154\u0026ndash;160\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJonsson U, Bohman H, von Knorring L, Olsson G, Paaren A, Von Knorring A-L (2011) Mental health outcome of long-term and episodic adolescent depression: 15-year follow-up of a community sample. J Affect Disord 130(3):395\u0026ndash;404\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKline RB (2023) Principles and practice of structural equation modeling. Guilford\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeatherdale ST, Brown KS, Carson V, Childs RA, Dubin JA, Elliott SJ, Faulkner G, Hammond D, Manske S, Sabiston CM (2014) The COMPASS study: a longitudinal hierarchical research platform for evaluating natural experiments related to changes in school-level programs, policies and built environment resources. BMC Public Health 14(1):1\u0026ndash;7\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeade AW, Craig SB (2012) Identifying careless responses in survey data. Psychol Methods 17(3):437\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMillsap RE, Kwok O-M (2004) Evaluating the impact of partial factorial invariance on selection in two populations. Psychol Methods 9(1):93\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMossman SA, Luft MJ, Schroeder HK, Varney ST, Fleck DE, Barzman DH, Gilman R, DelBello MP, Strawn JR (2017) The Generalized Anxiety Disorder 7-item (GAD-7) scale in adolescents with generalized anxiety disorder: signal detection and validation. Annals Clin psychiatry: official J Am Acad Clin Psychiatrists 29(4):227\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMotl RW, Dishman RK, Birnbaum AS, Lytle LA (2005) Longitudinal invariance of the Center for Epidemiologic Studies-Depression Scale among girls and boys in middle school. Educ Psychol Meas 65(1):90\u0026ndash;108\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMuthen LK, Muthen B, Muth\u0026eacute;n,M (2017) Mplus Version 8 User's Guide. Muthen \u0026amp; Muthen. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://books.google.ca/books?id=dgDlAQAACAAJ\u003c/span\u003e\u003cspan address=\"https://books.google.ca/books?id=dgDlAQAACAAJ\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRadloff LS (1977) The CES-D scale: A self-report depression scale for research in the general population. Appl Psychol Meas 1(3):385\u0026ndash;401\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRomano I, Ferro MA, Patte KA, Leatherdale ST (2021) Measurement invariance of the GAD-7 and CESD-R-10 among adolescents in Canada. J Pediatr Psychol\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchmitt N, Kuljanin G (2008) Measurement invariance: Review of practice and implications. Hum resource Manage Rev 18(4):210\u0026ndash;222\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSteinmetz H (2013) Analyzing observed composite differences across groups. Methodology\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVerhoeven M, Sawyer MG, Spence SH (2013) The factorial invariance of the CES-D during adolescence: are symptom profiles for depression stable across gender and time? J Adolesc 36(1):181\u0026ndash;190\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWidaman KF, Reise SP (1997) Exploring the measurement invariance of psychological instruments: Applications in the substance use domain\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang X, Kuchinke L, Woud ML, Velten J, Margraf J (2017) Survey method matters: Online/offline questionnaires and face-to-face or telephone interviews differ. Comput Hum Behav 71:172\u0026ndash;180\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[{"identity":"5c3cde28-023f-4d21-ba0e-a001b1e8fa96","identifier":"10.13039/501100000024","name":"Canadian Institutes of Health Research","awardNumber":"OOP-110788","order_by":0}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"University of Waterloo","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Depression, survey administration, COVID-19, youth, psychometric scale","lastPublishedDoi":"10.21203/rs.3.rs-5815737/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5815737/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDespite the prevalence of online surveys, research validating paper-based scale measurement properties for online use is lacking. We assessed the measurement invariance of the 10-item CESD-R-10 scale in adolescents transitioning from paper to online administration. We analyzed 2-year linked data from 2,831 Canadian secondary school students during 2019/20 and 2020/21. Using structural equation modeling, we examined measurement invariance across configural, metric, scalar, and strict stages due to the COVID-19 pandemic shift to online surveys. Baseline mean depression score was 8.2 (SD\u0026thinsp;=\u0026thinsp;5.9), with 33.7% (N\u0026thinsp;=\u0026thinsp;953) reporting clinically relevant symptoms. We found measurement invariance between paper and online administration. The results support measurement invariance in males, while partial scalar invariance for females when thresholds of items 6 (I felt fearful) and item 9 (I felt lonely) are freely estimated between the two administrations. Our findings demonstrate that the CESD-R-10 remains invariant even with a shift in data collection mode from paper to online surveys. This study supports the validity of conducting meaningful comparisons of depression symptoms when changing the mode of data collection in survey research among adolescent samples.\u003c/p\u003e","manuscriptTitle":"Measurement Invariance of the CESD-R-10 Among Adolescents over the Transition from Paper to Online Administration","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-14 07:19:59","doi":"10.21203/rs.3.rs-5815737/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c9e37f7e-529a-482b-960a-4dde0fb7a7f2","owner":[],"postedDate":"January 14th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-01-14T07:19:59+00:00","versionOfRecord":[],"versionCreatedAt":"2025-01-14 07:19:59","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5815737","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5815737","identity":"rs-5815737","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}