Contributed Poster Presentations: Social Statistics Section

Latent variable models (LVMs) are used across disciplines to investigate theories about the underlying structure of observed variables. Key to building LVMs is model fit evaluation. Fit is measured by how well the model reproduces the empirical covariance matrix (e.g., RMSEA) or by comparing the fit of multiple models (e.g., BIC). These traditional techniques are limited, ignoring the possibility of model misspecification, local misfit, concordance with prior research, and propensity of some models to fit most datasets well. This presentation offers a framework for assessing all aspects of LVM fit and synthesizing them into a holistic description to help researchers understand the relative appropriateness of different LVMs. This framework includes traditional global model fit (RMSEA), model fit comparison (BIC), as well as sensitivity analysis for model misspecification (ant colony optimization), local model fit, concordance with prior research, and fit propensity. Results across methods are synthesized, providing researchers with a more nuanced and potentially generalizable sense for how well a model fits the data. The presentation includes a complete example of the approach. 

The long-studied relationship between immigration and crime generally suggests a negative or null association, though few studies account for authorization status. Prior works have been limited by poor estimates of the number of unauthorized immigrants, and inadequate geographic resolution and coverage in crime data. In this study, we use a novel method of estimating the number of unauthorized immigrants in each US census tract, and examine the relationship between unauthorized immigration and annual tract-level crime rates from 10 varied police jurisdictions across the United States. To assess this relationship, we apply a linear model across census tracts, controlling for known correlates of crime. We find the association between unauthorized immigration and crime rates to be statistically insignificant across all jurisdictions and crime types, including drug, property, and violent crimes. This study can be used to further research into the association between authorization status and crime and inform public policy discussions. 

A wealth index is a measure to determine a households' aggregate living standard and socioeconomic position. In the absence of true information about household wealth, social sciences often construct proxy measures using information about household assets collected in population surveys. Such indices are used to position households in relative terms for studying differences by socioeconomic status. In literature, classical principal component analysis (PCA) has been extensively used to create different wealth indices. The aim of this study is to explore various modifications of the classical PCA approach, such as PCA with tetrachoric correlation matrix, and Sparse PCA with Pearson and tetrachoric correlation matrix to create household asset index using IPUMS International census microdata of developing countries. Spearman's rank correlation along with quintile based ranking of the households is used to compare the consistency of the indices derived using the various methods. 

Principal leadership is a multifaceted effort. The literature is barren on consistency of principal leadership across multiple policy domains. The issue is under-researched but important to leadership training and evaluation and critical to study teacher and student behaviors. The practical importance is operationalized in this paper. It will develop a multivariate, multilevel model that can measure consistency of principal leadership as perceived by teachers across multiple domains. The first level is a mathematical device to set up a multivariate environment. The second level is the teacher model in which teacher characteristics can be used to adjust their ratings of the multiple domains. The third level is the school model in which school characteristics can be used to adjust school (average) ratings of the multiple domains by teachers. A matrix is generated at the school level with variances for the multiple domains on the diagonal and covariances among the domains off the diagonal. This vector is distributed as multivariate normal. Correlations can be calculated from the variance-covariance matrix to measure consistency of principal leadership across the multiple domains. 

Music is known to have effects on the brain. With the prevalence of music streaming services today, it is essential to investigate the potential relationship between music and cognitive performance. However, while the effect of classical music on the brain, the so-called Mozart Effect, remains arguable, the effect of music tempo on memory recall performance is poorly understood. In this study, we evaluated the impact of music with various tempos on the memory recall ability among a cohort of high school students. Participants were first exposed to music of varying tempos or silence for 10 minutes and then engaged in a memory recall task. Our results show there is a statistically significant difference in the effects on memory performance between fast-tempo music and slow-tempo music (or no music at all) (p < 0.05). Compared with slow-tempo music, fast-tempo music also largely increased the participant's arousal level (p < 0.001). Taken together, our results indicate that faster-tempo music tends to enhance the listener's memory recall ability.