Identity theorists have emphasized the importance of integration across identity domains for psychosocial well-being. There remains little research, however, on associations across identity domains, group differences across identity profiles, and the joint association of multiple identity domains with academic outcomes. This dissertation includes two studies that address these limitations in the identity literature. Study 1, examined the ego-social identity profiles that emerged from ethnic identity exploration and commitment, American identity exploration and commitment, and ego identity integration and confusion among an ethnically diverse sample of emerging adults using latent profile analysis (N = 8,717). Results suggested that an eight-profile solution was the best fit for the data. The profiles demonstrated differences in identity status and salience across identity domains. Significant ethnic, sex, nativity, and age differences were identified in ego-social identity membership. Study 2 focused on the ego-social identity profiles that emerged from the same identity domains among biethnic college students of Latino and European American heritage (N = 401) and how these profiles differed as a function of preferred ethnic label. The association of ego-social identity profile with academic achievement and the moderation by university ethnic composition were examined. Results indicated that a two-profile solution was the best fit to the data in which one profile included participants with general identity achievement across identity domains and one profile included individuals who were approaching the identity formation process in each domain. Ego-social identity profile membership did not differ based on preferred ethnic label. Individuals who had a more integrated identity across domains had higher college grades. University ethnic composition did not significantly moderate this association. Taken together, these two studies highlight the intricacies of identity formation that are overlooked when integration across identity domains is not considered.

The effects of preventive interventions are found to be related to participants' responsiveness to the program, or the degree to which participants attend sessions, engage in the material, and use the program skills. The current study proposes a multi-dimensional method for measuring responsiveness to the Family Bereavement Program (FBP), a parenting-focused program to prevent mental health problems for children who experienced the death of a parent. It examines the relations between individual-level risk-factors and responsiveness to the program, as well as the relations between responsiveness and program outcomes. The sample consists of 90 caregivers and 135 children assigned to the intervention condition of an efficacy trial of the FBP. Caregivers' responsiveness to the 12-week program was measured using a number of indicators, including attendance, completion of weekly "homework" assignments, overall program skill use, perceived helpfulness of the program and program skills, and perceived group environment. Three underlying dimensions of responsiveness were identified: Skill Use, Program Liking, and Perceived Group Environment. Positive parenting and child externalizing problems at baseline were found to predict caregiver Skill Use. Skill Use and Perceived Group Environment predicted changes in caregiver grief and reports of child behavior problems at posttest and 11-month follow-up. Caregivers with better Skill Use had better positive parenting outcomes. Skill use mediated the relation between baseline positive parenting and improvements in positive parenting at 11-month follow-up.

The purpose of this study was to investigate the effect of complex structure on dimensionality assessment in compensatory and noncompensatory multidimensional item response models (MIRT) of assessment data using dimensionality assessment procedures based on conditional covariances (i.e., DETECT) and a factor analytical approach (i.e., NOHARM). The DETECT-based methods typically outperformed the NOHARM-based methods in both two- (2D) and three-dimensional (3D) compensatory MIRT conditions. The DETECT-based methods yielded high proportion correct, especially when correlations were .60 or smaller, data exhibited 30% or less complexity, and larger sample size. As the complexity increased and the sample size decreased, the performance typically diminished. As the complexity increased, it also became more difficult to label the resulting sets of items from DETECT in terms of the dimensions. DETECT was consistent in classification of simple items, but less consistent in classification of complex items. Out of the three NOHARM-based methods, χ2G/D and ALR generally outperformed RMSR. χ2G/D was more accurate when N = 500 and complexity levels were 30% or lower. As the number of items increased, ALR performance improved at correlation of .60 and 30% or less complexity. When the data followed a noncompensatory MIRT model, the NOHARM-based methods, specifically χ2G/D and ALR, were the most accurate of all five methods. The marginal proportions for labeling sets of items as dimension-like were typically low, suggesting that the methods generally failed to label two (three) sets of items as dimension-like in 2D (3D) noncompensatory situations. The DETECT-based methods were more consistent in classifying simple items across complexity levels, sample sizes, and correlations. However, as complexity and correlation levels increased the classification rates for all methods decreased. In most conditions, the DETECT-based methods classified complex items equally or more consistent than the NOHARM-based methods. In particular, as complexity, the number of items, and the true dimensionality increased, the DETECT-based methods were notably more consistent than any NOHARM-based method. Despite DETECT's consistency, when data follow a noncompensatory MIRT model, the NOHARM-based method should be preferred over the DETECT-based methods to assess dimensionality due to poor performance of DETECT in identifying the true dimensionality.

Item response theory (IRT) and related latent variable models represent modern psychometric theory, the successor to classical test theory in psychological assessment. While IRT has become prevalent in the assessment of ability and achievement, it has not been widely embraced by clinical psychologists. This appears due, in part, to psychometrists' use of unidimensional models despite evidence that psychiatric disorders are inherently multidimensional. The construct validity of unidimensional and multidimensional latent variable models was compared to evaluate the utility of modern psychometric theory in clinical assessment. Archival data consisting of 688 outpatients' presenting concerns, psychiatric diagnoses, and item level responses to the Brief Symptom Inventory (BSI) were extracted from files at a university mental health clinic. Confirmatory factor analyses revealed that models with oblique factors and/or item cross-loadings better represented the internal structure of the BSI in comparison to a strictly unidimensional model. The models were generally equivalent in their ability to account for variance in criterion-related validity variables; however, bifactor models demonstrated superior validity in differentiating between mood and anxiety disorder diagnoses. Multidimensional IRT analyses showed that the orthogonal bifactor model partitioned distinct, clinically relevant sources of item variance. Similar results were also achieved through multivariate prediction with an oblique simple structure model. Receiver operating characteristic curves confirmed improved sensitivity and specificity through multidimensional models of psychopathology. Clinical researchers are encouraged to consider these and other comprehensive models of psychological distress.