Do Prison Administrative and Survey Data Sources Tell the Same Story? A Multitrait, Multimethod Examination With Application to Gangs

Abstract

Measurement is critical to advancing theory and research in criminology. Yet, criminologists are often forced to rely on data sources that are not intended to be used for research or collected in contexts where subjects may have incentives to misreport. This is particularly true of data in institutional corrections research. This study leveraged novel data to examine correspondence in key measures found in prison administrative and survey data on 802 prisoners in Texas, focusing on the measurement properties of an important group rarely studied in prison: gang members. We observed high rates of correspondence between data sources for gang membership (82%) and the gang with which they were affiliated (86%). A multilevel test of item correspondence demonstrated that the measures of gang membership performed as well or better than more episodic measures and worse than more durable measures common in corrections research. A multitrait, multimethod matrix revealed that gang membership, and nearly all other measures, satisfied the principles of validity. Finally, there were no differences in correspondence rates between gang and nongang members for nearly every measure, regardless of the method to measure gang membership. Prison administrative and survey data generally tell the same story, which is promising for institutional corrections research, particularly research focused on gangs.