Analyzing repeated-game economics experiments: robust standard errors for panel data with serial correlation

The purpose of this study is to provide guidance to those who analyze data from repeated-game experiments. In particular, I propose the use of heteroskedasticity-autocorrelation consistent (HAC) covariance estimators for panel data, which allows researchers to conduct hypothesis tests without having to place structure on the heteroskedasticity and/or serial correlation likely present in econometric models. Through Monte Carlo experiments I explore the properties of three panel HAC covariance estimators within a linear regression framework, including a new HAC covariance estimator proposed in this study, for a range of cross-section (