Effects of VariablesUsing Panel Data: A Review of Techniques

One motivation for using panel data is to determine the relativeeffects of two variables which are known to be related. Thisarticle reviews the four techniques which have been used mostoften to address this problem. The relative merits of the fourLazarsfeld's 16 Fold Table, Coleman's Four State ContinuousTime Markov Process, Cross Lagged Panel Correlation, and PathAnalysis and the particular problems associated with each arereviewed. The first three methods are shown to yield similarresults, the fourth to do so under some assumptions.