Testing the Assumptions of Pooled Time Series Cross-Sectional Regression Analysis Through ARIMA and Interrupted Time Series Modeling: Beck and Katz Revisited

Literature: Three prominent models of time series regression analysis exist in statistical analysis for the social sciences: pooled time series cross-sectional regression analysis, ARIMA time series analysis and interrupted time series analysis. Purpose: The purpose of this manuscript is to measure the assumptions of pooled time series cross-sectional regression analysis in formal modeling by testing a Mississippi casino data set, overtime. Data: The data set used in this manuscript is a compilation of social indicators, over a twelve-year period, measuring the impact of casino revenues on per pupil expenditures for education. Methods: To validate the findings of the pooled time series cross-sectional regression analysis, both ARIMA time series analysis, and interrupted time series analysis were used to manipulate the casino data set. Results: All three time series models reported similar results in regards to the casino data set. Conclusion: This research suggests that the Beck and Katz’s (1995) argument supporting pooled time series cross-sectional regression analysis, as a valid model of regression modeling, is substantiated. Limitations: Due to the limited time period covered by the data set, future studies of similar magnitude should incorporate lengthier time spans for more reliable time series analysis (Thirty years preferably).