Brief report: Testing measurement invariance and differences in self-concept between adolescents with and without physical illness or developmental disability

Self-concept in Adolescents with Physical-Mental Comorbidity

Objective

Little is known about self-concept in adolescents with physical-mental comorbidity. This study investigated whether physical-mental comorbidity was associated with self-concept in adolescents and examined if adolescent age or sex moderated the association between physical-mental comorbidity and self-concept.

Methods

Study data were obtained from the Multimorbidity in Youth across the Life-course (MY LIFE), an ongoing Canadian study of adolescents with chronic physical illness who were recruited from outpatient clinics at a pediatric hospital. A total of 116 adolescents aged ≥ 10 years provided self-reports on key measures.

Results

Adolescents with comorbidity (n = 48) had lower self-concept scores on the Self-Determination Questionnaire (SDQ; d = 0.62) and Self-Perception Profile for Children (SPPC; d = 0.53) vs. adolescents without comorbidity (n = 68). An age × comorbidity status interaction was found and age-stratified models were computed to investigate this moderating effect of age. Amongst older adolescents, comorbidity was associated with lower SDQ (B = -2.55, p < .001), but this association was not found among younger adolescents (B = -0.29, p = .680). A similar effect was found for SPPC among older (B = -0.48, p = .001) and younger adolescents (B = 0.03, p = .842). Adolescent sex was not found to be a moderator.

Conclusions

Physical-mental comorbidity in adolescence was associated with lower self-concept and this association was moderated by age-differences between adolescents with vs. without comorbidity were greater for older adolescents and were clinically relevant. Opportunities to support positive self-perceptions for adolescents with comorbidity are warranted, especially when planning the transition from pediatric to adult health services.

Factor structure and measurement invariance of the Chinese version of the COVID-19 Phobia Scale in depressive symptoms sample during COVID-19 closure: An exploratory structural equation modeling approach

The COVID-19 Phobia Scale is an instrument for measuring the phobia of coronavirus. It has a stable four-factor structure and good reliability and validity in other countries and regions. In order to expand related research, this study aims to test the reliability and validity of the COVID-19 Phobia Scale in Chinese adolescents with depressive symptoms. The C19P-SC was translated into Chinese by the method of forward and back translation and tested in 1933 Chinese adolescents with depressive symptoms. Confirmatory factor analysis (CFA) and exploratory structural equation modeling (ESEM) were used to test and compare the four-factor model of the C19P-SC. Then we tested the measurement invariance of the C19P-SC across gender and time. Finally, the reliability was measured with the McDonald's omega coefficients. Consistent with previous studies, the C19P-SC showed a stable four-factor structure. The results showed that ESEM was better than CFA and more reasonable. In addition, the results of multi-group ESEM showed that the C19P-SC met the strict invariance at male and female and partial longitudinal strict invariance. The Mcdonald's omega coefficients of the C19P-SC total scale and each subscale reached the expected acceptable level. In short, the reliability and validity index of C19P-SC has reached an acceptable level, and the measurement invariance of different genders and different time points was established, but the cross-factor phenomenon of individual items was abnormal, and a further revision and testing are still needed.

Child Sexual Abuse, Self-esteem, and Delinquent Behaviors During Adolescence: The Moderating Role of Gender

To reflect the complex phenomena of child sexual abuse (CSA), studies should examine possible gender specificities and explore potentially explanatory mechanisms. The current study aimed to test the moderating effect of gender in the mediated relationship between CSA, self-esteem, and delinquency during adolescence. A moderated mediation model was tested among a representative sample of 8,194 high school students (57.8% girls and 42.2% boys) age 14 to 18 in the province of Quebec in Canada. Results showed that self-esteem has an indirect effect on the relationship between CSA and delinquency. Gender (being a boy) was associated with a higher level of self-esteem and an increased risk of delinquent behaviors. Among victims of CSA, boys reported lower levels of self-esteem than girls, which was associated with an increased risk of displaying delinquent behaviors. Self-esteem may be an important target of intervention for sexually abused youth, especially for boys. Focusing on promoting positive self-esteem may also reduce the risk for male adolescents struggling with the deleterious consequences of delinquency.

Measurement invariance of the Positive Gains Scale in families of children with and without disabilities

BACKGROUND

Despite the high frequency of case-control studies in the developmental disability literature, there is a paucity of research establishing the measurement equivalence of instruments used, and particularly those relating to positive perceptions and experiences in family disability research.

AIMS

The present study sought to establish measurement invariance for the Positive Gains Scale (PGS) across 1219 mothers of children with developmental disabilities, 234 mothers of children with spina bifida/hydrocephalus, and 157 mothers of children without disabilities.

METHODS AND PROCEDURES

A three-step test for measurement invariance across the three groups was conducted using Multigroup Confirmatory Factor Analysis.

OUTCOMES AND RESULTS

Loadings between the three groups were invariant, suggesting the criteria to assume metric invariance was met. However, the assumption of scalar invariance was not met, suggesting that item intercepts differed between the three groups.

CONCLUSIONS AND IMPLICATIONS

Our findings suggest that the PGS cannot be meaningfully used to compare outcomes between mothers of children with developmental disabilities and other mothers. These findings may have wider implications for research utilising well-being measures to make comparisons with carers of children with developmental disabilities.

Psychometric properties of the 10-item Connor-Davidson Resilience Scale (CD-RISC-10) in Chinese undergraduates and depressive patients

BACKGROUDS

The 10-item Connor-Davidson Resilience Scale (CD-RISC-10) is a widely self-report questionnaire to assess resilience in different populations, including adolescents, elderly individuals and psychiatric patients. Considering the application of the CD-RISC-10 in depression patients, the present study aimed to evaluate the psychometric properties, especially the measurement invariance, of the CD-RISC-10 in depressive patient sample.

METHODS

A total of 2230 undergraduates from Hunan Province and 293 patients with major depressive disorder (MDD) from psychological clinics participated in our study. Confirmatory factor analysis (CFA) was conducted to evaluate the single-factor model of the CD-RISC-10 and the measurement equivalence of the CD-RISC-10 across the clinical (MDD) and non-clinical (undergraduate) samples, as well as gender invariance in the non-clinical sample.

RESULTS

The findings implied that the CD-RISC-10 showed good reliability and validity, including favorable internal consistency and criterion-related validity. The CFA results showed that the strong model fit for a unidimensional structure of the CD-RISC-10 was supported in the clinical (eg., CFI >0.910, TLI >0.900, RMSEA < 0.080) and non-clinical (eg., CFI >0.950, TLI >0.940, RMSEA < 0.060) groups, respectively. Moreover, the scalar invariance of the CD-RISC-10 was supported across the clinical and non-clinical samples (eg., △CFI < 0.009, △TLI < 0.005, ΔRMSEA < 0.003). Similarly, the strict gender invariance was also established in the male (n = 1035) and female (n = 1195) undergraduate samples (eg., △CFI < 0.008, △TLI < 0.004, ΔRMSEA < 0.002).

LIMITATIONS

The inclusion of Chinese individuals only and the absence of depressive patients with comorbidities were our major limitations.

CONCLUSIONS

Overall, these psychometric findings not only supported the stability and validity of the CD-RISC-10 for application with different samples in resilience study, but also indicated that the CD-RISC-10 could be an effective instrument for research in depression.