Bayesian Analysis of State Voter Registration Database Integrity

How do we ensure a statewide voter registration database’s accuracy and integrity, especially when the database depends on aggregating decentralized, sub-state data with different list maintenance practices? We develop a Bayesian multivariate multilevel model to account for correlated patterns of change over time in multiple response variables, and label statewide anomalies using deviations from model predictions. We apply our model to California’s 22 million registered voters, using 25 snapshots from the 2020 presidential election. We estimate countywide change rates for multiple response variables such as changes in voter’s partisan affiliation and jointly model these changes. The model outperforms a simple interquartile range (IQR) detection when tested with synthetic data. This is a proof-of-concept that demonstrates the utility of the Bayesian methodology, as despite the heterogeneity in list maintenance practices, a principled, statistical approach is useful. At the county level, the total numbers of anomalies are positively correlated with the average election cost per registered voter between 2017 and 2019. Given the recent efforts to modernize and secure voter list maintenance procedures in the For the People Act of 2021, we argue that checking whether counties or municipalities are behaving similarly at the state level is also an essential step in ensuring electoral integrity.

Bayesian Analysis of State Voter Registration Database Integrity

How do we ensure a statewide voter registration database's accuracy and integrity, especially when the database depends on aggregating decentralized, sub-state data with different list maintenance practices? We develop a Bayesian multivariate multilevel model to account for correlated patterns of change over time in multiple response variables, and label statewide anomalies using deviations from model predictions. We apply our model to California's 22 million registered voters, using 25 snapshots from the 2020 presidential election. We estimate countywide change rates for multiple response variables such as changes in voter's partisan affiliation and jointly model these changes. The model outperforms a simple interquartile range (IQR) detection when tested with synthetic data. This is a proof-of-concept that demonstrates the utility of the Bayesian methodology, as despite the heterogeneity in list maintenance practices, a principled, statistical approach is useful. At the county level, the total numbers of anomalies are positively correlated with the average election cost per registered voter between 2017--2019. Given the recent efforts to modernize and secure voter list maintenance procedures in the For the People Act of 2021, we argue that checking whether counties or municipalities are behaving similarly at the state level is also an essential step in ensuring electoral integrity.

Relationships between teacher influence in managerial and instruction-related decision-making, job satisfaction, and professional commitment: A multivariate multilevel model

Research shows teacher influence in school decision-making is related to improving individual job satisfaction and professional commitment. However, few empirical studies investigate how different domains of school decision-making may have a distinct relationship with both individual teacher job satisfaction and professional commitment. Using the Schools and Staffing Survey 2011–2012 and the National Teacher and Principal Survey 2015–2016 which are national samples of U.S. primary and secondary schools, this study examines teacher influence in managerial and instruction domains of school decision-making and whether both are associated with job satisfaction and professional commitment. Results showed that teacher influence in the instruction domain has a statistically stronger relationship with teacher professional commitment than job satisfaction. In contrast, teacher influence in the managerial domain of school decision-making has a statistically stronger relationship with teacher job satisfaction than professional commitment. The results have policy implications for involving teachers in school decision-making and improving their overall job satisfaction.

A Multivariate Multilevel Analysis of the Risk Factors Associated with Anthropometric Indices in Iranian mid-adolescents

Background The present study was conducted to assess some specific factors related to anthropometric indices in a representative sample of Iranian mid-adolescents. It also aimed to investigate the relationship between three anthropometric indices using a multivariate multilevel approach.Methods This study was conducted among 2538 students (1286 boys) aged 14 - 20 years old, who were randomly selected among 16 high schools by multi-stage random sampling procedure from 4 education districts of Shiraz, Iran. Data on demographic characteristics and anthropometric measures were collected. Anthropometric indices were estimated by the calibrated tools. A multivariate multilevel model was used to assess the predictor variables associated with obesity measures of the triceps (TST), abdominal (AST), and subscapular (SST) skinfold thickness.Results In this study, the prevalence of overweight and obesity was estimated to be 10.2 and 5.1%, respectively. Hierarchical models showed that, the outcome variables were correlated at the child and district levels. Positive associations were observed between the sex, family history of obesity, and moderate and high Socio-Economic Status (SES) with three anthropometric indices. Moreover, there were positive associations between moderate to vigorous physical activities with AST and SST, but they were not statistically significant at a significance level of 0.05.Conclusions Our findings revealed that, TST, AST, and SST could be useful indices for measuring the adiposity in mid-adolescents. In addition, the effect of the district level on the predictor variables highlights the important role of the environmental factors on the childhood obesity.

Organizational Culture and Job Demands and Resources: Their Impact on Employees’ Wellbeing in a Multivariate Multilevel Model

(1) The present study aims to explore the impact of job demands and resources (JDR), personal resources, and the organizational culture on workers’ wellbeing and health. (2) A cross-sectional survey of Spanish workers in small and medium-sized enterprises (SMEs) was conducted with a sample of 1599 workers from 154 SMEs. A multivariate multilevel analysis was performed to analyze the different relationships. (3) In light of the results obtained, we observed that job demands were negatively associated with workers’ health, while job resources were positively correlated to workers’ health and wellbeing. Secondly, the different types of identification at work are positively related to job satisfaction and organizational citizenship behaviors (OCBs), but the intensity of this association differs with the form of identification. Finally, at the organization level, the dimensions of organizational culture are related differently to employees’ job satisfaction, OCBs, and health; (4) these results confirm the role of organizational culture and its association with desirable outcomes, allowing us to expand the JDR model.

Nonlinear Strain Effects on Delinquent Behavior and Depressive Symptoms

Objectives: Few studies have examined the potential nonlinear effects of strain on offending and other outcomes, even though the initial development of general strain theory (GST) recognized their feasibility. This study was designed to investigate these nonlinear effects on delinquent behavior and depressive symptoms, as well as the potential moderating effects of family relations. Methods: Eight years of longitudinal data from the Family Health Study ( n = 840) were used to develop a multivariate multilevel model that examined the nonlinear effects of strain—in the form of stressful life events—on delinquent behavior and depressive symptoms. Results: The empirical results suggested that strain has an exponential effect on the delinquency when family relations are poor and a diminishing effect on depressive symptoms when family relations are strong. Conclusions: The findings of this study lend support to the argument that strain has nonlinear effects on delinquency, crime, and depressive symptoms. This suggests that research should consider nonlinear effects in more detail when assessing of GST. Moreover, good family relations can serve to protect youth against the untoward effects of strain and thus may provide a practical focus for intervention efforts.

Exploring Off-Season Content and Interaction on Facebook: A Comparison of U.S. Professional Sport Leagues

Research on social-media use in sport should be expanded to include analyses of content popularity and comparisons across leagues. This study used content analysis and a multivariate multilevel model to compare content type and interaction across U.S. professional sport leagues. Results indicated that teams in the National Football League had the most comments, teams in the Major League Baseball had the most shares, and teams in the National Basketball Association had the most likes. Content coded as player and personnel promotion, which included behind-the-scenes content and human-interest stories, received the most interaction. Sport marketers can use this information to drive content strategy. However, content designed to encourage interaction is still posted less often than most other types of content. These results suggest that marketers in sport may be using Facebook to build relationships by connecting fans personally with players, but not by encouraging interaction or 2-way conversation.