Construct validity and dimensionality of the measure of criminal social identity using data drawn from American, Pakistani, and Polish inmates

Purpose: The purpose of this paper is to test the construct validity and dimensionality of the measure of criminal social identity (MCSI) within both a combined sample of American, Pakistani, and Polish inmates, as well as examined as individual country samples. Design/methodology/approach: Adopting a cross-sectional survey design, the opportunistic sample consisted of offenders incarcerated in three different countries; 351 inmates from Poland, 501 from the USA, and 319 from Pakistan (combined data set n=1,171), with inmates completing anonymous, self-administered, paper-and-pencil questionnaires. Traditional confirmatory factor analysis, along with confirmatory bi-factor modelling, was used in order to examine the fit of four different models of criminal social identity (CSI). Findings: Results revealed that data were best explained by a three-factor model of CSI (cognitive centrality, in-group ties, and in-group affect) within both combined and individual offender samples. Composite reliability indicated that the three factors were measured with very good reliability. Research limitations/implications: Validation of the MCSI within the large cross-cultural combined prison sample provides substantial support for the measure’s reliability and utility across diverse offender samples. Consideration of low factor loadings of items one and three for the Pakistan data set and item two for the US data set, leads the researchers to outline possible recommendations that these questions be reworded and additional items be added. Originality/value: This is the first study to validate MCSI cross-culturally and specifically utilising a western prison sample, consisting of male and female offenders.