Amanda Goodrick

We propose an extension of a previously proposed method for lead-lag analysis of multivariate time series to include the analysis of spatial correlations. We applied the extended spatial and temporal method to CIRIGHTS, a large global human rights dataset, in order to determine the most influential and most influenced indicators of human rights, freedoms, and atrocities over time. We consider four target countries, each from a different continent. The previously proposed method used a weighted directed network with several lags of each variable as nodes and with edges weighted by transfer entropy. In this study, that method is extended to the spatial multivariate time series by investigating geographical correlations between each target country and its surrounding region to determine which variables best represent activity between the target country and the surrounding region. The nodes in this undirected network are variables in the target country and variables in the surrounding region, with edges weighted by mutual information. Eigenvector centrality is used to determine which variables best represent activity between the target country and the surrounding region. Key findings indicate that occupational and worker rights are the most influential and most influenced in the target countries over time and by region, and laws tend to influence future activity.

We propose a method of analyzing multivariate time series data that investigates lead-lag relationships among economic indicators during the COVID-19 era with a weighted directed network of lagged variables. The analysis includes a stock index, average unemployment, and several variables that are used to calculate inflation. Three complex networks are created, with these variables and several lags of each as the network nodes. Network edges are weighted based on three relationship metrics: correlation, mutual information, and transfer entropy. In each network, nodes are merged, and edges are aggregated to simplify the weighted directed graph. Pagerank is used to determine the most influential and the most influenced node over the time period. Results were reasonably robust within each network, but they were heavily dependent on the choice of metric.

Because it utilizes a combination of high impact practices—a collaborative project, service learning, and undergraduate research—the use of a community-engaged framework within an undergraduate introductory statistics course can improve educational outcomes for distance learners. During the Fall of 2020, because the usual community-engaged approach of having students partner with a community organization to collect and analyze data was not feasible, we used a recently released dataset on coping mechanisms during the COVID-19 pandemic. Students participated in an analysis of those data in order to inform policy-making within the university regarding support for students during the pandemic. Similar outcomes in terms of improved understanding of and attitudes toward statistics were achieved through this approach as had been achieved through community-engaged projects in previous semesters. Moving to a distance learning format within the pandemic has required rethinking how these components of community-engaged learning can be preserved. This case study shows that similar learning outcomes can be achieved despite the limitations the pandemic places on learning.