Document Type
Journal Article
Publication Date
Winter 12-2020
Department
Agricultural Science
College/School
Hutson School of Agriculture
Abstract
The Chow test is a standard method to test for differences in regression response across groups. In some cases, the groups being tested are composed of a time series of cross sections. If the individual units within the groups have systematic differences, the Chow test is compromised: individual and group effects become confounded. This can cause rejections in the absence of the group effect of interest. We illustrate the problem with Monte Carlo analyses, and propose an alternative bootstrap-like testing procedure that helps eliminate excessive Type I errors.
Recommended Citation
Binkley, James K. and Young, Jeffrey, "The Chow Test with Time Series-Cross Section Data" (2020). Faculty & Staff Research and Creative Activity. 184.
https://digitalcommons.murraystate.edu/faculty/184
Included in
Agribusiness Commons, Agricultural and Resource Economics Commons, Econometrics Commons, Health Economics Commons, Industrial Organization Commons, Regional Economics Commons