<p>The power of positive attitudes toward science is that they influence science achievement by reinforcing higher performance. Interestingly, there continue to be gender disparities in attitudes toward science across many countries. Males generally have more positive attitudes toward science than females. Although most research related to attitudes toward science have been based on the Trends in International Mathematics and Science Study (TIMSS) Student Questionnaire, there remains a dearth of evidence validating the TIMSS science attitudes items and measurement equivalence across genders. </p><p>The goals of this research were as follows: (1) to build support for the structural validity of the TIMSS items, and (2) to investigate whether the instrument measures the same latent construct (attitudes toward science) across genders. The present study followed two steps of statistical analyses. As a first step, two modeling methods (confirmatory factor analysis and exploratory structural equation modeling) were conducted to identify the best-fitting model for the instrument. Second, after determining the model of choice, we tested several nested invariance models progressively. </p><p>This study found (1) the latent factor structure of the TIMSS items and (2) strong measurement invariance across genders. This result indicated that the instrument is well designed by the <i>a priori</i>specification and measures the same latent variable for both female and male students. This study provides support for the multidimensional approach to measuring science attitudes and shows the flexibility of ESEM over CFA by demonstrating that the ESEM approach provided better representation of the underlying factor structure. </p>

Researchers are often interested in comparing brain activity between experimental contexts. Event-related potentials (ERPs) are a common electrophysiological measure of brain activity that is time-locked to an event (e.g., a stimulus presented to the participant). A variety of decomposition methods has been used for ERP data among them temporal exploratory factor analysis (EFA). Essentially, temporal EFA decomposes the ERP waveform into a set of latent factors where the factor loadings reflect the time courses of the latent factors, and the amplitudes are represented by the factor scores. An important methodological concern is to ensure the estimates of the condition effects are unbiased and the term variance misallocation has been introduced in reference to the case of biased estimates. The aim of the present thesis was to explore how exploratory factor analytic methods can be made less prone to variance misallocation. These efforts resulted in a series of three publications in which variance misallocation in EFA was described as a consequence of the properties of ERP data, ESEM was proposed as an extension of EFA that acknowledges the structure of ERP data sets, and regularized estimation was suggested as an alternative to simple structure rotation with desirable properties. The presence of multiple sources of (co-)variance, the factor scoring step, and high temporal overlap of the factors were identified as major causes of variance misallocation in EFA for ERP data. It was shown that ESEM is capable of separating the (co-)variance sources and that it avoids biases due to factor scoring. Further, regularized estimation was shown to be a suitable alternative for factor rotation that is able to recover factor loading patterns in which only a subset of the variables follow a simple structure. Based on these results, regSEMs and ESEMs with ERP-specific rotation have been proposed as promising extensions of the EFA approach that might be less prone to variance misallocation. Future research should provide a direct comparison of regSEM and ESEM, and conduct simulation studies with more physiologically motivated data generation algorithms.

Heroin addiction is inclined to arouse fear, rejection and discriminatory behavior among the general public. Evidence shows that the public perceives heroin as harmful and addictive. Heroin is ranked as the most stigmatized condition. While there is robust literature on mental illness stigma, there is limited research concerning addiction-related stigma. There are very few standardized stigma measures related to perceptions toward persons addicted to heroin. The overall aim of the dissertation was to validate an attribution measurement model toward persons addicted to heroin and to determine its psychometric properties. The dissertation’s study employed an adapted 7-factor measurement model (Corrigan et al., 2002) to examine stigmatizing perceptions towards persons addicted to heroin. This is the first study to systematically evaluate model fit by implementing Exploratory Structural Equation Modeling (ESEM). A total of 657 Sociology students were analyzed over four stages: questionnaire review by expert panel, pilot-test, validation and replication. The study tested multiple incremental models and successfully determined that the results met multiple goodness-of-fit indices. Through ESEM, Sociology-Social Control students supported the hypothesis that the adapted 7-factor attribution measurement model would fit data. The model included: Personal Responsibility, Pity, Anger, Helping Behavior, Dangerousness, Fear and Social Distance factors. Adequate power and sample size was demonstrated to support acceptance of the null hypothesis. In addition to conducting Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA), ESEM tested the psychometric properties of the attribution measurement model. Implementing maximum likelihood extraction with oblique geomin rotation using Mplus software, the Sociology-Social Control students’ validation and replication datasets showed an excellent model fit to the data. Results confirmed support for the superiority of the ESEM solution. The ESEM attribution measurement model fit better than the CFA model. Compared to the ESEM model, elevated factor correlations found in the CFA model were caused by the exclusion of meaningful cross-loadings. Strong psychometric properties for the ESEM attribution model were evidenced, with good internal consistency and excellent test-retest reliability. The factor structure was replicable across the two groups of Sociology-Social Control students. Adequate ESEM incremental and convergent validity was supported by the simultaneous examination of the Social Distance scale and the Personal Consequences of Criminal Stigma measures with the measurement model. In the replication sample, familiarity demonstrated less stigmatizing perceptions than the SOC313 Course. Our findings highlight marked differences between the Sociology-Social Control students and the general population’s perceptions of heroin addicts. The Sociology-Social Control students are not afraid of persons addicted to heroin, nor do they hold them responsible for their condition. To conclude, the study provides newly validated measures with adequate reliability to allow investigators to assess other students’ level of addiction stigma. It is anticipated that the dissertation’s study will lead to further comparative psychometric testing with healthcare students that are directly involved with the care and treatment of persons addicted to heroin to provide a better understanding of the factorial structure of the attribution measurement model. Longitudinal data is also needed to examine our model and how levels of perceptions change over time.

Social-emotional learning (SEL) measures assessing social-emotional learning and character development across a broad array of constructs have been developed but lack construct validity. Determining the efficacy of educational interventions requires structurally valid measures which are generalizable across settings, gender, and time. Utilizing recent factor analytic methods, the present study extends validity literature for SEL measures by investigating the structural validity and generalizability of the Social-Emotional and Character Development Scale (SECDS) with a large sample of children from schools in Belize (n = 1877, ages 8 to13). The SECDS exhibited structural and generalizability evidence of construct validity when examined under exploratory structural equation modeling (ESEM). While a higher order confirmatory factor structure with six secondary factors provided acceptable fit, the ESEM six-factor structure provided both substantive and methodological advantages. The ESEM structural model situates the SECDS into the larger body of SEL literature while also exhibiting generalizability evidence over both gender and time.