Assessment最新文献

The Child and Adolescent PsychProfiler version 5 (CAPP v5) is a comprehensive multi‑informant screening measure encompassing 17 symptom scales that map onto 14 Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5) disorders. The self‑report form (CAPP‑SRF) has not previously undergone a comprehensive psychometric evaluation. The objective of the study is to analyze the internal structure (Independent Clusters Model of Confirmatory Factor Analysis [ICM‑CFA]), reliability (α, ω), and validity evidence (discriminant, convergent, criterion‑related) of the CAPP‑SRF. Study 1 examined the 17‑factor model within a community sample of 790 adolescents (M = 14.48 years). Study 2 evaluated convergent, criterion‑related, and discriminant validity in a clinic‑referred sample of 173 adolescents (M = 14.50 years) utilizing the Conners 3‑SR, Beck Youth Inventories, Second Edition (BYI‑2), Wechsler Intelligence Scale for Children-Fifth Edition (WISC‑V), and Wechsler Individual Achievement Test, Third Edition (WIAT‑III). Independent‑samples t tests compared CAPP‑SRF means across samples. The ICM‑CFA analysis confirmed the 17‑factor structure (χ²/df = 3.02; standardized root mean squared error [SRMR] = .076). Scale reliability was acceptable (ω = .79-.89). Clinic participants scored significantly higher than community participants on 15 of the 17 scales (all p < .001; d = .55-1.20), supporting criterion validity. Convergent and discriminant patterns with external measures were as hypothesized (|r| = .32-.68; R² = .10-.46). The CAPP‑SRF demonstrates robust psychometric properties and complements the parent‑ and teacher‑report forms as an effective adolescent self‑report screener for common DSM‑5 disorders.

The Mental Help Seeking Attitudes Scale (MHSAS) is commonly used in psychological research. This nine-item unidimensional scale was designed to measure how favorable or unfavorable respondents' attitudes are toward seeking help from a mental health professional and was originally validated using a primarily White sample. To address the potential limitations of using this scale cross-culturally, we recruited participants who identified as Asian American (n = 161), Black American (n = 259), and Latine American (n = 259) to take the MHSAS, and then we ran confirmatory factor analyses. Our samples also all consisted of individuals with chronic pain. The original measure validation demonstrated excellent overall model fit; however, all three of our non-White samples had only adequate overall model fit and factor loading values. An exploratory bifactor analysis still confirmed the unidimensional structure of the scale. These findings suggest that the MHSAS should be used cautiously in cross-cultural contexts with racially minoritized groups experiencing chronic pain.

The Meaning in Life Questionnaire (MLQ) assesses two distinct dimensions of meaning in life: presence of meaning and search for meaning. The MLQ is the most widely used instrument for measuring meaning in life, yet there is a limited variety of validity evidence on the originally proposed two-factor confirmatory factor analysis (CFA) solution. In this light, the present research examined, across five studies (total N = 3,205), several aspects of the MLQ's validity and tested cross-gender and cross-national measurement invariance. We also examined the usefulness of the exploratory structural equation model (ESEM) of the MLQ as an alternative to the standard CFA model. The results obtained provide evidence for: (a) the validity (structural, convergent, concurrent, and incremental) of the MLQ ESEM factors; (b) full scalar invariance of the MLQ ESEM model across gender and partial measurement invariance across four countries; and (c) similar cross-national relationships between MLQ ESEM factors and measures of depression and life satisfaction. The present research provides support for the value of applying the ESEM framework in overcoming limitations of the CFA model when examining evidence on the MLQ's validity.

This study examined the item- and scale-level functioning of the Social Appearance Anxiety Scale (SAAS) as well as differential functioning by gender using an item response theory (IRT) analysis. SAAS data collected from 840 college students were analyzed. A graded response model was used to analyze the 16 items comprising the SAAS. The measure was found to be unidimensional in its factor structure, and every item demonstrated high to very high ability to differentiate respondents varying in levels of the underlying trait (i.e., appearance concerns). In addition, we found evidence of differential item functioning (DIF) by gender for four items, corresponding to small effect sizes. Two of these items were related to internal experiences of appearance concerns (e.g., nervousness and discomfort when a flaw is noticed by others) that were more likely to be endorsed by women, and two of the items were related to external evaluative experiences related to appearance (e.g., missing opportunities and life being more difficult) that were more likely to be endorsed by men. Overall, the IRT and DIF results suggest that the SAAS effectively identifies appearance concerns among individuals with low to very high appearance concerns.

The present study applied taxometric analyses to the Washington Assessment of Risks and Needs of Students (WARNS)-an instrument designed to assess multiple domains of functioning related to justice system involvement arising from school disengagement-a trajectory referred to as the school to prison pipeline. Previous taxometric studies of constructs related to juvenile justice system involvement found dimensional rather than taxonic (dichotomous) latent structures. Participants were 5008 students from 89 Washington school districts who completed the WARNS as part of standard educational practices. The results were uniformly consistent with a dimensional latent structure. Also supporting a dimensional latent structure, dichotomized WARNS scores were significantly less strongly associated with student arrests, school suspensions, and school skip days than continuous WARNS scores. These findings support the dimensionality of risk and needs and have implications for assessments undertaken to improve school and social outcomes for at-risk youth.

Ecological Momentary Assessment using smartphone technology (smart EMA) has grown substantially over the last decade. However, little is known about the factors associated with completion rates in populations who have a higher likelihood of cognitive impairment. A systematic review of Smart EMA studies in populations who have a higher likelihood of cognitive impairment was carried out (PROSPERO; ref no CRD42022375829). Smartphone EMA studies in neurological, neurodevelopmental and neurogenetic conditions were included. Six databases were searched, and bias was assessed using Egger's test. Completion rates and moderators were analyzed using meta-regression. Fifty-five cohorts were included with 18 cohorts reporting confirmed cognitive impairment. In the overall cohort, the completion rate was 74.4% and EMA protocol characteristics moderated completion rates. Participants with cognitive impairment had significantly lower completion rates compared with those without (p = .021). There were no significant moderators in the cognitive impairment group. Limitations included significant methodological issues in reporting of completion rates, sample characteristics, and associations with completion and dropout rates. These findings conclude that smart EMA is feasible for people with cognitive impairment. Future research should focus on the efficacy of using smart EMA within populations with cognitive impairment to develop an appropriate methodological evidence base.

This study examined associations of compliance rate with the reliability and convergent validity of intoxication and negative affect assessments in experience sampling method (ESM) data in three samples (Veterans, Sexual Minority Men, and College Students). Convergent validity was operationalized as within-person associations between daily aggregates of random in situ assessments and retrospective daily assessments or transdermal alcohol assessments. Measures with lower ICC require more assessments for a reliable aggregate (e.g., daily mean). In this regard, the number of completed assessments and intraclass correlation (ICC), rather than compliance with the protocol per se, determines reliability. Although convergent validity was correlated with compliance rate, the relatively weak associations reflect that there are individuals with excellent compliance yet poor convergent validity as well as individuals with poor compliance and excellent convergent validity. The pattern of results does not show a clear threshold for compliance (e.g., 80%) that differentiates good versus poor validity.

Decades of research show a clear link between family factors and psychopathology. Family functioning varies across cultures, suggesting potential cultural differences in the association between family factors and psychopathology. In addition, assessing family functioning generally involves tools not systematically validated for diverse cultural backgrounds. Using the Adolescent Brain Cognitive Development^SM data (N = 11,864), this study found: (a) full scalar invariance was tenable for the Children's Reports of Parental Behavior Inventory (CRPBI) and Prodromal Questionnaire-Brief Child Version (PQ-BC) across race/ethnicity, but not for the Family Environment Scale (FES) and Parental Monitoring Survey (PMQ); (b) the CRPBI and PMQ were significantly associated with the PQ-BC, and (c) all three family scales had equivalent relations with the PQ-BC across groups. This highlights the importance of evaluating scales for measurement invariance across race/ethnicity. Results also help to connect specific family factors to the etiology of psychosis risk among U.S. children and adolescents.

Psychological test scores are commonly used in high-stakes settings to classify individuals. While measurement invariance across groups is necessary for valid and meaningful inferences of group differences, full measurement invariance rarely holds in practice. The classification accuracy analysis framework aims to quantify the degree and practical impact of noninvariance. However, how to best navigate the next steps remains unclear, and methods devised to account for noninvariance at the group level may be insufficient when the goal is classification. Furthermore, deleting a biased item may improve fairness but negatively affect performance, and replacing the test can be costly. We propose item-level effect size indices that allow test users to make more informed decisions by quantifying the impact of deleting (or retaining) an item on test performance and fairness, provide an illustrative example, and introduce unbiasr, an R package implementing the proposed methods.

Some recent studies have revived the approach of investigating extreme levels of self-reported depressive symptoms as indicative of gross exaggeration. While scores above 40 on the Beck Depression Inventory-Second Edition (BDI-II) were discussed as indicating exaggerated symptom claims, different cut scores for identifying noncredible responding are now being discussed. A consecutive sample of 242 patients referred for forensic psychological assessment (mean age: 46.0 years, 47.7% women) with full data sets on the BDI-II and the Structured Inventory of Malingered Symptomatology (SIMS) were assessed. Of all patients, 13.2% scored above 40 and BDI-II scores correlated with SIMS total scores at .62. For different SIMS cutoffs (>14, >16, >19, >23) used as criterion standard, optimal cut scores for the BDI-II were computed. When specificity was set at a minimum of 90%, sensitivity estimates were below 50% for all four SIMS levels. Extreme scores on the BDI-II should raise concern about the credibility of self-reported depressive symptom load. Diagnosis as well as severity estimates should not be based primarily on self-report instruments. To avoid significant risks of bias, the development of reliable cut scores for BDI-II elevations should be based on more studies with samples from diverse contexts.