Maximum Likelihood Confirmatory Factor Analysis Research Articles

Overview and confirmatory and exploratory factor analysis of AI literacy scale

Comprehensive concepts of AI literacy (AIL) and valid measures are essential for research (e.g., intervention studies) and practice (e.g., personnel selection/development) alike. To date, several scales have been published, sharing standard features but differing in some aspects. We first aim to briefly overview instruments identified from unsystematic literature research in February 2023. We identified four scales and one collection of items. We describe the instruments and compare them. We identified common themes and overlaps in the instruments and developmental procedure. We also found differences regarding scale development procedures and latent dimensions. Following this literature research, we came to the conclusion that the literature on AI literacy measurement was fragmented, and little effort was undertaken to integrate different AI literacy conceptualizations. The second focus of this study is to test the factorial structures of existing AIL measurement instruments and identify latent dimensions of AIL across all instruments. We used robust maximum-likelihood confirmatory factor analysis to test factorial structures in a joint survey of all AIL items in an English-speaking online sample (N=219). We found general support for all instruments' factorial structures with minor deviations from the original factorial structures for some of the instruments. In a second analysis step, to address the issue of fragmented research on AI literacy conceptualization and measurement, we used principal axis exploratory factor analysis with oblique rotation to identify latent dimensions across all items. We found four correlated latent dimensions of AIL, which were mostly interpretable as the abilities to use and interact with AI, to design/program AI (incl. in-depth technical knowledge), to perform complex cognitive operations regarding AI (e.g., ethical considerations), and a common factor for the abilities to detect AI/differentiate between AI and humans and manage persuasive influences of AI (i.e., persuasion literacy). Our findings sort the multitude of AIL instruments and reveal four latent core dimensions of AIL. Thus, they contribute importantly to the conceptual understanding of AIL that has been fragmented so far.

Influence of General Health on Quality of Life by the Caregiver and Burden Care as a Mediator: Bayesian Estimator Structural Equation Analysis

Objectives: To test structural equation models of General Health on quality of life by the caregiver and burden care as a mediator. Methods: The quality of life scale was administered to 297 caregivers of persons with schizophrenia. The models were examined using maximum likelihood confirmatory factor analysis (ML-CFA) and Bayesian structural equation modeling (BSEM). The path analysis with direct and indirect effects was discussed in the model. Results: The results reveal that the models fit well for evidence-based. The BIC 1412.998 < 3161.387, DIC 1365.247 < 3094.673, and WAIC 1370.802 < 3094.651 were the lowest BIC of the models and displayed a simple and parsimonious factor-loading pattern. Conclusions: Overall, both of the scales demonstrated excellent measurement properties supporting its relevance to comprehensively measure caregivers' impact and burden on persons with schizophrenia. Psychiatric and mental health care teams and clinical researchers could use the quality of life scale to generate strategies to support the family caregiver and measure the multidisciplinary treatment team interventions' outcomes.

Investigating discrimination in the workplace. Translation and validation of the Everyday Discrimination Scale for nursing staff in Germany

BackgroundThe Everyday Discrimination Scale (EDS) is a frequently used questionnaire in the field of health and social psychology that aims to explore perceptions of discrimination, especially instances of injustice related to various diversity characteristics. No adaptation to health care staff exists. The present study translates and adapts the EDS to nursing staff in Germany and examines its reliability and factorial validity as well as its measurement equivalence between men and women and different age groups.MethodsThe study was based on an online survey conducted among health care staff of two hospitals and two inpatient care facilities in Germany. The EDS was translated using a forward-backward translation approach. Direct maximum likelihood confirmatory factor analysis (CFA) was conducted to examine the factorial validity of the adapted EDS. Differential item functioning (DIF) related to age and sex was investigated by means of multiple indicators, multiple causes (MIMIC) models.ResultsData on 302 individuals was available, of whom 237 (78.5%) were women. The most commonly employed one-factor, 8-item baseline model of the adapted EDS showed a poor fit (RMSEA = 0.149; CFI = 0.812; TLI = 0.737; SRMR = 0.072). The model fit improved considerably after including three error covariances between items 1 and 2, items 4 and 5, and items 7 and 8 (RMSEA = 0.066; CFI = 0.969; TLI = 0.949; SRMR = 0.036). Item 4 showed DIF related to sex and age, item 6 showed DIF related to age. DIF was moderate in size and did not bias the comparison between men and women or between younger and older employees.ConclusionsThe EDS can be considered a valid instrument for the assessment of discrimination experiences among nursing staff. Given that the questionnaire, similar to other EDS adaptations, may be prone to DIF and also considering that some error covariances need to be parameterized, latent variable modelling should be used for the analysis of the questionnaire.

Development and primary validation of the School Health Assessment Tool for Primary Schools (SHAT-PS).

BackgroundSchool health programs need to target all aspects of physical, psychological, and social well-being. Using a slightly modified version of the COSMIN Risk of Bias checklist, we developed and conducted the first validation of the School Health Assessment Tool for Primary Schools (SHAT-PS).MethodThe exploratory sequential mixed method was used in this study. In the first phase, scientific databases were systematically searched to find school health models and instruments and 65 interviews were conducted with school stakeholders. The Colaizzi’s method was used to code the qualitative data into themes. Then, a pool of items was created for each theme, rechecked by psychometric experts and then validated for content (i.e., relevance, clarity, and comprehensiveness) by psychometric experts and individuals of the target population (i.e., school personnel). In the second phase, classical test theory was utilized to analyze the validity and reliability of the resulting items from phase 1 among 400 individuals working at primary schools.ResultsThe coding of the interviews resulted in ten themes that we labeled based on the theoretical literature: school health policies, community connections, health education, physical activity, health services, nutrition, psychological services, physical environment, equipment and facilities, and school staff’s health. The items created for each theme ended up in an initial pool of 76 items. In the final stage of phase 1, 69 items remained after the content validity assessment by experts and school personnel. In phase 2, the SHAT-PS items were tested using maximum likelihood exploratory factor analysis and confirmatory factor analysis. Of the 69 items from phase 1, 22 items were removed due to low factor loadings. The results showed that the 8-factor model was the best solution (chi-square/df = 2.41, CFI = .98, TLI = .97, RMSEA = .06). The discriminant and convergent validity of the SHAT-PS were evaluated as satisfactory and the scale had high internal consistency (Cronbach’s alpha for all subscales > .93). The test-retest reliability was satisfactory—the intraclass correlation coefficient pooled was .95 (99% CI [.91–.98]). Moreover, the standard error of measurement resulted in an SEM pooled equal to 4.4. No discrepancy was found between subgroups of gender and subgroups of staffs’ positions at schools.ConclusionThe SHAT-PS is a valid and reliable tool that may facilitate school staff, stakeholders and researchers to evaluate the presence of the factors that promote health at primary schools. Nevertheless, in the process of validation, many of the items related to staff’s health were eliminated due to poor factor loadings. Obviously, staff health is an important factor in the measurement of school health. Hence, we recommend that the validity and reliability of the SHAT-PS in other cultures should be done using the original 76-item version.

Bayesian and Maximum-Likelihood Modeling and Higher-Level Scores of Interpersonal Problems With Circumplex Structure.

Difficulties in interpersonal behavior are often measured by the circumplex-based Inventory of Interpersonal Problems. Its eight scales can be represented by a three-factor structure with two circumplex factors, Dominance and Love, and a general problem factor, Distress. Bayesian confirmatory factor analysis is well-suited to evaluate the higher-level structure of interpersonal problems because circumplex loading priors allow for data-driven adjustments and a more flexible investigation of the ideal circumplex pattern than conventional maximum likelihood confirmatory factor analysis. Using a non-clinical sample from an online questionnaire study (N = 822), we replicated the three-factor structure of the IIP by maximum likelihood and Bayesian confirmatory factor analysis and found great proximity of the Bayesian loadings to perfect circumplexity. We found additional support for the validity of the three-factor model of the IIP by including external criteria-Agreeableness, Extraversion, and Neuroticism from the Big Five and subclinical grandiose narcissism-in the analysis. We also investigated higher-level scores for Dominance, Love, and Distress using traditional regression factor scores and weighted sum scores. We found excellent reliability (with Rtt ≥ 0.90) for Dominance, Love, and Distress for the two scoring methods. We found high congruence of the higher-level scores with the underlying factors and good circumplex properties of the scoring models. The correlational pattern with the external measures was in line with theoretical expectations and similar to the results from the factor analysis. We encourage the use of Bayesian modeling when dealing with circumplex structure and recommend the use of higher-level scores for interpersonal problems as parsimonious, reliable, and valid measures.

Psychometric Validation of the Indonesian Version of the Fear of COVID-19 Scale: Personality Traits Predict the Fear of COVID-19

The coronavirus disease-2019 (COVID-19) pandemic is a global health crisis that has generated fear and negative psychological consequences. The present study evaluated the validity and factor structure of Fear of COVID-19 Scale (FCV-19S) among a sample from the general Indonesian population. The English version of the FCV-19S was translated and back-translated into Indonesian language, followed by a pilot study. Using convenience sampling method, a total of 728 participants completed an online survey distributed on various social media platforms. The survey included the FCV-19S, personality traits (neuroticism and extraversion), Positive and Negative Affect Scale (PANAS), Generalized Anxiety Disorder Scale (GAD-7), and the nine-item Patient Health Questionnaire (PHQ-9). The Indonesian FCV-19S had very good internal consistency (Cronbach’s alpha and McDonald’s omega) and composite reliability (alpha = 0.88, omega = .86, composite reliability = .87). Maximum likelihood confirmatory factor analysis (CFA) was conducted to test construct validity (χ2/df = 2.51, CFI = .984, SRMR = .028, PCLOSE = .15 > .05, RMSEA = .06, 90% CI [.03, .09]). As for criterion-related validity, the FCV-19S score positively correlated with the score on PHQ-9, GAD-7, negative affect, and neuroticism and negatively correlated with extraversion. Negative affect was identified as the most important predictor of the fear of COVID-19. Personality traits also predicted the fear of COVID-19. The findings provide evidence that the FCV-19S is a reliable and valid instrument for assessing fear generated by COVID-19 among a healthy Indonesian-speaking population.

Psychometric Validation of the Persian Version of the COVID-19-Related Psychological Distress Scale and Association with COVID-19 Fear, COVID-19 Anxiety, Optimism, and Lack of Resilience.

The outbreak of the coronavirus disease-2019 (COVID-19) has resulted in a global health crisis. The COVID-19 pandemic has caused psychological distress, both in infected and uninfected individuals. The present study evaluated the validity and factor structure of the COVID-19-Related Psychological Distress Scale (CORPDS) among the general public of the Persian-speaking population. The original version of the CORPDS was translated and back-translated into Persian, followed by a pilot study. A total sample (n = 623) completed an online survey including the CORPDS, Fear of COVID-19 Scale (FCV-19S), Coronavirus Anxiety Scale (CAS), Kessler Psychological Distress Scale (K10), Life Orientation Test-Revised (LOT-R), and Brief Resilience Scale (BRS). The Persian CORPDS had very good internal consistency and moderate test-retest reliability after 4 weeks. Maximum likelihood confirmatory factor analysis (CFA) was conducted to test construct validity (χ2/df = 2.39, CFI = 0.95, SRMR = 0.046, PCLOSE = 0.67 > 0.05, RMSEA = 0.047, 90% CI [0.038, 0.056]). Measurement invariance was performed across gender, including configural invariance, metric invariance, scalar invariance, and error variance invariance, and yielded further support for the two-factor structure of the CORPDS. The CORPDS correlated with the score on the K10 (r = 0.46, p < 0.01, 95% CI [0.43, 0.48]), CAS (r = 0.43, p < 0.01, 95% CI [0.37, 0.45]), FCV-19S (r = 0.29, p < 0.01, 95% CI [0.27, 0.32]), LOT-R (r = − 0.19, p < 0.01, 95% CI [− 0.15, − 0.24]) and BRS (r = − 0.56, p < 0.01, 95% CI [− 0.50, − 0.61]). Resilience was associated with lower psychological distress (β = − 0.54, SE = 0.05, p < 0.001). The findings provide evidence that CORPDS is a reliable and valid instrument for assessing psychological distress generated by COVID-19 among a healthy Persian-speaking population.

BSSI-3Spoken language, reading–writing forms: Validity and reliability study

This study was conducted to determine whether the ‘Spoken Language’ and ‘Reading–Writing’ forms of the Basic School Skills Inventory-3 Edition had validity and reliability characteristics acceptable for Turkish children aged 48–72 months old. The sample of the study included 244 children selected through the simple random sampling method. Exploratory factor analysis (principal components analysis) was conductedto determine the construct validity of the tests. The fit index was tested with maximum likelihood confirmatory factor analysis. Reliability of the test was ensured using internal consistency coefficient and split-half test. Cronbach’s alpha reliability coefficients of the forms were calculated to be 0.95 and 0.97. Analysed for its adaptation into Turkish, the BSSI-3 Spoken Language and Reading–Writing forms were a valid and reliable assessment tool for Turkish children aged 48–72 months old. Keywords: Spoken language, reading–writing, validity and reliability

Psychometric Validation of the Persian Version of the Problem Gambling Severity Index

The Problem Gambling Severity Index (PGSI) has been the most frequently used instrument for prevalence studies for problem gambling in recent years. To date, there have been very few gambling studies among Persian populations. The objective of the present study was to investigate the reliability and validity of the Persian version of the PGSI. The original version of the PGSI was translated and back-translated into Persian, followed by a pilot study. A sample of Iranian online gamblers (n = 858) was recruited utilizing social media platforms. Results showed that the Persian PGSI had excellent internal consistency (Cronbach’s alpha coefficient = 0.90; composite reliability = 0.91) and moderate test-retest reliability after 4–6 weeks using intraclass coefficient (0.41 with 95% CI [.34, .49]). The maximum likelihood confirmatory factor analysis for one-factor PGSI model fitted well (χ2/df = 2.81, CFI = .974, SRMR = .028, PCLOSE > .05, RMSEA = .057, 90% CI [.037, .077]). As for criterion-related validity, the Persian PGSI score correlated with the score on the Hospital Anxiety Depression Scale, which assesses depression (r = .54, p < .01) and anxiety (r = .40, p < .01). The Persian PGSI is a reliable and valid instrument for assessing the prevalence of problem gambling among Persian populations.

Development and Scoring of a Survey on Public Health Accreditation Capacity.

Public health accreditation is a 7-step process that starts with a period of preapplication during which a health department assesses its readiness for accreditation. However, no tools with established reliability and validity that quantitatively measure a local health department's (LHD's) capacity for accreditation are available to complete this initial step. We developed and validated a survey to measure accreditation capacity for LHDs. From January through April 2015, we administered a cross-sectional electronic survey instrument with 15 questions that tapped into domains of capacity for public health accreditation. We analyzed and grouped responses by using a confirmatory maximum likelihood factor analysis with oblique rotations. We assessed reliability by using Cronbach α, and we assessed validity by comparing responses with previously established instruments. We administered the survey to 174 LHD directors in Colorado, Kansas, and Nebraska, 153 (88%) of whom responded. The factor analysis produced a 3-factor model of accreditation capacity, suggesting that accreditation capacity depends on 3 distinct latent constructs: support for accreditation, preparation, and planning and approach. The model had good scale reliability (average Cronbach α = 0.7) and validity (average factor correlation = 0.43). The survey developed and scored in this analysis can be used by LHDs to inform the feasibility of initiating the time-intensive and costly process of accreditation.

Measurement Invariance of Three Versions of the Pathological Narcissism Inventory across Gender-Matched Italian Adolescent High School and Young Adult University Students

To assess the measurement invariance of the three versions of the Pathological Narcissism Inventory (PNI) across adolescent and young adult participants, 678 Italian adolescent high school students (M = 16.83 years old, SD = 1.85 years) and 678 adult university students (M = 23.99 years old, SD = 2.60 years) matched on gender, civil status, and geographic area, were administered the Italian translation of the original 52-item PNI. The PNI, the Brief-Pathological Narcissism Inventory (B-PNI), and Super Brief-Pathological Narcissism Inventory (SB-PNI) scales all showed adequate reliabilities in both adolescent and young adult samples. In both samples, robust maximum likelihood confirmatory factor analysis indicated that covariation among the PNI/B-PNI first-order scales, as well as among the SB-PNI items, could be adequately explained by two-correlated factors operationalizing narcissistic grandiosity (NG) and narcissistic vulnerability (NV), respectively. Multiple-group confirmatory factor analyses supported scalar invariance for both B-PNI and SB-PNI NG and NV factors, and metric invariance for the PNI NG and NV factors.

Enabling and leveraging ambidexterity: influence of strategic orientations and knowledge stock

PurposeThis paper aims to develop and empirically test a framework articulating the effects of strategic orientations (entrepreneurial orientation and market orientation) on leveraging ambidexterity. Further, the paper examines the moderating effects of knowledge stock (market knowledge and technological knowledge) on the relationship between ambidexterity and firm performance to gain additional insights into how ambidexterity can be leveraged in an organization.Design/methodology/approachData were obtained from CEOs (or equivalent members of the top management team) of 234 firms. The adequacy and psychometric properties of all measures were evaluated and purified using a maximum likelihood confirmatory factor analysis (CFA), and the hypotheses were tested using ordinary least squares (OLS). A number of post hoc tests were conducted to develop a nuanced understanding of proposed effects.FindingsWhile both strategic orientations enhance an organization’s ability to be ambidextrous, results show that some types of knowledge stocks facilitate, whereas other types hinder the influence of ambidexterity on firm performance.Research limitations/implicationsBoth strategic orientations enable ambidexterity; however, technological knowledge stock impedes the effect of ambidexterity on firm performance, while market knowledge stock enhances this relation. Cross-sectional nature of the study imposes limitations on causal inferences.Practical implicationsDifferent strategic orientations provide organizations with a cluster of knowledge acquisition and utilization capabilities that enable ambidexterity. However, organizations should be wary of indiscriminate accumulation of knowledge stocks – while certain types enhance the effect of ambidexterity, others may create competency traps or core rigidities and inhibit the effect of ambidexterity.Originality/valueThis study integrates related, yet hitherto fragmented, research streams to demonstrate the interconnectedness between strategic orientations, ambidexterity and existing knowledge stock. Several theoretical and managerial implications are identified.

Further Insights Into the German Version of the Multidimensional Assessment of Interoceptive Awareness (MAIA)

Abstract. Interoception is defined as an iterative process that refers to receiving, accessing, appraising, and responding to body sensations. Recently, following an extensive process of development, Mehling and colleagues (2012) proposed a new instrument, the Multidimensional Assessment of Interoceptive Awareness (MAIA), which captures these different aspects of interoception with eight subscales. The aim of this study was to reexamine the dimensionality of the MAIA by applying maximum likelihood confirmatory factor analysis (ML-CFA), exploratory structural equation modeling (ESEM), and Bayesian structural equation modeling (BSEM). ML-CFA, ESEM, and BSEM were examined in a sample of 320 German adults. ML-CFA showed a poor fit to the data. ESEM yielded a better fit and contained numerous significant cross-loadings, of which one was substantial (≥ .30). The BSEM model with approximate zero informative priors yielded an excellent fit and confirmed the substantial cross-loading found in ESEM. The study demonstrates that ESEM and BSEM are flexible techniques that can be used to improve our understanding of multidimensional constructs. In addition, BSEM can be seen as less exploratory than ESEM and it might also be used to overcome potential limitations of ESEM with regard to more complex models relative to the sample size.

Construct validity and diagnostic accuracy of the Italian translation of the 18-Item World Health Organization Adult ADHD Self-Report Scale (ASRS-18) Italian translation in a sample of community-dwelling adolescents.

Aim of this prospective study was to evaluate the factor structure, internal consistency, and diagnostic efficacy of the Italian translation of the Adult ADHD Self-Report Scale (ASRS-18; Kessler et al., 2005, 2007) in a sample of community-dwelling sample of adolescents. Three hundred eight Italian adolescents attending professional high schools were administered the ASRS-18; the adolescence ADHD module of the MINI interview was administered to obtain ADHD criterion diagnoses. Robust maximum likelihood confirmatory factor analysis results identified a bifactor model of the ASRS-18 items as the best-fitting model, RMSEA = 0.029, p > 0.99, TLI = 0.93, CFI = 0.95. Further analyses showed that 77.9% of the ASRS-18 reliable (ω = 0.78) score variance was due to the total score. In our study the ASRS-18 proved to be able to effectively differentiate adolescents who received a MINI ADHD diagnosis (n = 80) from adolescents who did not receive a MINI ADHD diagnosis (n = 163), area under the curve = 0.80, 95% confidence interval = 0.74, 0.86. Our data suggest that the ASRS-18 may represent an effective self-report measure to screen for ADHD in community-dwelling adolescents, at least in its Italian translation.

A Psychometric Analysis of the Student Leadership Competencies Inventory

The Student Leadership Competencies Inventory consists of eight scales, each corresponding to its relevant leadership construct within the Student Leadership Competencies framework (Seemiller, 2013). Due to the increasing use of the framework and associated inventory in leadership development programs in higher education, we conducted a thorough analysis of the psychometric properties within each scale. Specifically, using a national dataset of university student responses, we analyzed internal reliability statistics, and conducted exploratory factor analysis with varimax rotation and maximum likelihood confirmatory factor analysis for each of the eight scales. Our results suggested that all scales, overall, possess sufficient strength to be considered valid measures of the leadership constructs within the Inventory, with some notably high co-variances between certain sub-scale factors in several scales.

The psychometric properties of a modified client-centred rehabilitation questionnaire in an Australian population

Purpose: To identify the face validity of the Client-Centred Rehabilitation Questionnaire (CCRQ) and to determine the internal consistency reliability and factorial validity of a modified CCRQ.Materials and methods: This study was conducted in 2 phases. Phase 1 consisted of 5 focus groups to examine the face validity of the CCRQ, resulting in the development of a modified CCRQ. Phase 2 consisted of a multi-site cross sectional survey, involving 408 rehabilitation inpatients, to examine the internal consistency reliability and factorial validity of the modified CCRQ. Chronbach’s coefficient alpha, composite reliability coefficients, and single factor congeneric models with maximum likelihood confirmatory factor analysis were used.Results: Based on feedback from the focus groups the CCRQ was modified with the word ‘rehabilitation’ replacing ’program’ throughout. The three negatively worded items had poor item-to-total correlations of <0.3. Removing these items resulted in subscale alphas of 0.74–0.86 and composite reliability coefficients of 0.66–0.87. Six of the seven sub-scales had good model fit and the other one had moderate fit following removal of the negatively worded item.Conclusions: This study supports the underlying structure and internal consistency of the modified CCRQ.Implications for RehabilitationPerson-centredness is an important characteristic of effective rehabilitation service delivery that warrants measurement.A modified Client-Centered Rehabilitation Questionnaire has been found to have good face validity, internal consistency reliability and construct validity in an Australian sample of inpatient rehabilitation patients.Use of a modified Client-Centered Rehabilitation Questionnaire incorporating the 7 sub-scales in the original Client-Centered Rehabilitation Questionnaire is supported for use in inpatient rehabilitation.Both sub-scale and item level responses to the modified Client-Centered Rehabilitation Questionnaire provide detailed feedback to rehabilitation service providers looking for opportunities to make their services more person-centred.

Differential Item Functioning Analysis Of The Physical Activity Enjoyment Scale

BACKGROUND: Enjoyment is a commonly reported motivation for engaging in physical activity and may be a key mediator of exercise promoting interventions for teens. Given the burden of physical inactivity in the US, identification of correlates and potential targets of intervention to promote activity are needed. The Physical Activity Enjoyment Scale (PACES) represents a promising tool to evaluate patterns of physical activity and sedentary behavior among teens. Continued validation for use with multi-ethnic teens is needed, particularly among Hispanic teens who may be characterized by risk factors associated with lower engagement in physical activity and higher sedentary behavior. METHOD: A school-based sample of 9th Grade Hispanic (N = 1605; mean age = 14.6 years, SD = .41) and Non-Hispanic White (N = 544; mean age = 14.6 years, SD = .39) students completed in-classroom self-report measures of physical activity (PA), sedentary behavior (SED), and the PACES. Full Information Maximum likelihood confirmatory factor analysis and item response models (IRM) of the Positive and Negative PACES subscales assessed construct validity and Differential Item Functioning (DIF) of the 14-item PACES among Latino and Non-Latino adolescents. RESULTS: Non-parametric IRM suggested modifications to the response format for negatively worded subscale of the PACES. A two-factor model reflecting both positive and negative subscales provided better fit than a unidimensional model (X2= 4755.0, p<0.001). Loadings on primary factors ranged from 0.87-0.94 for positive and 0.78-0.96 for negative experiences subscales. Internal consistency estimates were 0.95 and 0.90 for positive and negative scales, respectively. DIF analyses supported measurement invariance and construct validity of both subscales across Hispanic and Non-Hispanic White respondents. PACES-Positive and PACES-Negative maintained significant correlations with levels of physical activity (r=0.38 & r=-0.31, respectively) and moderate to vigorous physical activity (r=0.31 & r=-0.19, respectively). CONCLUSIONS: We found support for the construct validity, concurrent validity and consistency of measurement of the 14-item PACES provides for use in comparing positive and negative experiences of exercise among both Non-Hispanic White and Hispanic teens.

Confirmatory Factor Analysis of the Patient Version of the Working Alliance Inventory–Short Form Revised

The working alliance concerns the quality of collaboration between patient and therapist in psychotherapy. One of the most widely used scales for measuring the working alliance is the Working Alliance Inventory (WAI). For the patient-rated version, the short form developed by Hatcher and Gillaspy (WAI-SR) has shown the best psychometric properties. In two confirmatory factor analyses of the WAI-SR, approximate fit indices were within commonly accepted norms, but the likelihood ratio chi-square test showed significant ill-fit. The present study used Bayesian structural equations modeling with zero mean and small variance priors to test the factor structure of the WAI-SR in three different samples (one American and two Swedish; N = 235, 634, and 234). Results indicated that maximum likelihood confirmatory factor analysis showed poor model fit because of the assumption of exactly zero residual correlations. When residual correlations were estimated using small variance priors, model fit was excellent. A two-factor model had the best psychometric properties. Strong measurement invariance was shown between the two Swedish samples and weak factorial invariance between the Swedish and American samples. The most important limitation concerns the limited knowledge on when the assumption of residual correlations being small enough to be considered trivial is violated.

Strengthening the assessment of factorial invariance across population subgroups: a commentary on Varni et al. (2013)

This article provides a commentary in response to "Varni et al. (Qual Life Res. doi: 10.1007/s11136-013-0370-4 , 2013)." The commentary argues that the approximate model fit indexes commonly used in maximum-likelihood confirmatory factor analysis and factorial invariance testing are seriously flawed, as they overlook potentially serious model misspecifications that could bias parameter estimates and compromise inference. Flexible and convenient Bayesian estimation approaches are presented that can substantially aid in: (1) resolving commonly encountered specification errors in confirmatory factor models and (2) locating specific measurement parameters that are non-invariant across population subgroups. It is recommended that these methods should be more widely adopted for evaluating the factorial invariance of patient-reported outcome measures and other types of instruments.

Factor analyses of the Hospital Anxiety and Depression Scale: a Bayesian structural equation modeling approach

The latent structure of the Hospital Anxiety and Depression Scale (HADS) has caused inconsistent results in the literature. The HADS is frequently analyzed via maximum likelihood confirmatory factor analysis (ML-CFA). However, the overly restrictive assumption of exact zero cross-loadings and residual correlations in ML-CFA can lead to poor model fits and distorted factor structures. This study applied Bayesian structural equation modeling (BSEM) to evaluate the latent structure of the HADS. Three a priori models, the two-factor, three-factor, and bifactor models, were investigated in a Chinese community sample (N=312) and clinical sample (N=198) using ML-CFA and BSEM. BSEM specified approximate zero cross-loadings and residual correlations through the use of zero-mean, small-variance informative priors. The model comparison was based on the Bayesian information criterion (BIC). Using ML-CFA, none of the three models provided an adequate fit for either sample. The BSEM two-factor model with approximate zero cross-loadings and residual correlations fitted both samples well with the lowest BIC of the three models and displayed a simple and parsimonious factor-loading pattern. The study demonstrated that the two-factor structure fitted the HADS well, suggesting its usefulness in assessing the symptoms of anxiety and depression in clinical practice. BSEM is a sophisticated and flexible statistical technique that better reflects substantive theories and locates the source of model misfit. Future use of BSEM is recommended to evaluate the latent structure of other psychological instruments.