A PC PROGRAM FOR COMPUTING MCMAHANS TRACKING INDEXES FROM ONE-SAMPLE LONGITUDINAL DATA SETS

Tracking refers to regular, predictable growth behavior. Most definiti ons of the term are related to the idea that tracking occurs to the ex tent that individuals' repeated measurements maintain their relative p ositions in the distribution of the measurement whose growth is being monitored as that distribution changes over time. In this paper we des cribe four indices of tracking, all due to McMahan (Biometrics, 37 (19 81) 447-455), which may be used to measure the extent to which a given longitudinal sample exhibits this 'regular behavior,' and we provide a program, written in GAUSS386i, for performing the associated computa tions. An example is given, and the values of McMahan's indices are co mpared to several alternative measures of tracking.