Back

Presentation

Cook’s distance measures for panel-data models

David Vincent

8 September 2022

Session

Influential observations in regression analysis are data points whose deletion has a large impact on the estimated coefficients.

The usual diagnostics for assessing the influence of each data point are designed for least-squares regression and independent observations and are not appropriate when estimating panel-data models.

 

The purpose of this presentation is to describe a new command, cooksd2, that extends the traditional Cook’s (1977) distance measure to determine the influence of each data point when applying the fixed-, random-, and between-effects regression estimators. The approach is based on the framework developed by Christensen, Pearson, and Johnson (1992) and also reports the influence of an entire subject or group of data points following the methods described by Banerjee and Frees (1997).

References:

Cook, R. D. (1977). Detection of influential observation in linear regression. Technometrics 19: 15–18.

Banerjee, M., and E. W. Frees. (1997). Influence diagnostics for linear longitudinal models. Journal of the American Statistical Association 92: 999 1005.

Christensen, R., L. M. Pearson, and W. Johnson. 1992. Case-deletion diagnostics for mixed models. Technometrics 34: 38–45.

Speaker

David Vincent