Combined Approach to Multi-Informant Data Using Latent Factors and Latent Classes: Trifactor Mixture Model

Dealing with missing data in multi-informant studies: A comparison of approaches

Multi-informant studies are popular in social and behavioral science. However, their data analyses are challenging because data from different informants carry both shared and unique information and are often incomplete. Using Monte Carlo Simulation, the current study compares three approaches that can be used to analyze incomplete multi-informant data when there is a distinction between reference and nonreference informants. These approaches include a two-method measurement model for planned missing data (2MM-PMD), treating nonreference informants' reports as auxiliary variables with the full-information maximum likelihood method or multiple imputation, and listwise deletion. The result suggests that 2MM-PMD, when correctly specified and data are missing at random, has the best overall performance among the examined approaches regarding point estimates, type I error rates, and statistical power. In addition, it is also more robust to data that are not missing at random.

Multi-informant Assessment of Internalizing Concerns: Rater Concordance and Implications for Decision-Making

Rising rates of mental health challenges among youths have become a significant concern following the COVID 19 pandemic. Although strong evidence supports the implementation of universal screening as a preventative approach to address unmet mental health concerns, the research is less clear surrounding the use of such data in decision-making processes when significant discrepancies between informants (e.g., students and teachers) exist. The purpose of the study was twofold. First, the study aimed to determine the degree of rater concordance between teachers and students on students' internalizing concerns. The second objective was to determine whether concordance on internalizing behaviors differs across ages/grades and if this differentially impacts distal (i.e., academic) outcomes. Results indicated that teachers and students demonstrated limited agreement on ratings of internalizing behaviors. However, when students and teachers agreed, higher and more positive emotional behaviors were linked to higher reading/math performance. Furthermore, patterns of informant dis/agreement and relationships between internalizing concerns and academic outcomes were similar across grade levels. Implications and areas for future research are discussed.

Can Faking Be Measured With Dedicated Validity Scales? Within-Subject Trifactor Mixture Modeling Applied to BIDR Responses

A sample of 516 participants responded to the Balanced Inventory of Desirable Responding (BIDR) under answer honest and instructed faking conditions in a within-subjects design. We analyze these data with a novel application of trifactor modeling that models the two substantive factors measured by the BIDR-self-deceptive enhancement (SDE) and impression management (IM), condition-related common factors, and item-specific factors. The model permits examination of invariance and change within subjects across conditions. Participants were able to significantly increase their SDE and IM in the instructed faking condition relative to the honest response condition. Mixture modeling confirmed the existence of a theoretical two-class solution comprised of approximately two thirds of "compliers" and one third of "noncompliers." Factor scores had good determinacy and correlations with observed scores were near unity for continuous scoring, supporting observed score interpretations of BIDR scales in high-stakes settings. Correlations were somewhat lower for the dichotomous scoring protocol. Overall, results show that the BIDR scales function similarly as measures of socially desirable functioning in low- and high-stakes conditions. We discuss conditions under which we expect these results will and will not generalize to other validity scales.