Propensity To Support Prison Gangs Summary

The article “Propensity to Support Prison Gangs: Its Relationship to Gang Membership. Victimization, Aggression and Other Disruptive Behaviours”, by Jane Ireland and Christina Powers seek to depict the relationship between the tendency to support gang membership and other problematic behaviours within prison. The researchers situate the article within prison, group dynamics and gang literature, which suggest that they are interested in utilizing all three areas to identify alternative factors that influence gang membership and associated behaviours. The researchers conclude that the propensity to support prison gangs was associated with aggression and other disruptive behaviours, gang membership and victimization. The authors employ a descriptive

research design, which aims at uncovering possible relationships within the findings. The following critique will demonstrate how the researchers inability to clearly define their theoretical constructs as well as the use of reactive measures significantly reduces our confidence in the generalizability of their conclusions. In other words, had less reactive measures been used and had the theoretical constructs been defined differently, the findings may have been different.
To begin, the researchers put into question the external validity of their findings due to their inability to adequately define their theoretical constructs.

Although the researchers take into account the complexities of defining gangs and gang membership, they do not clearly define the construct. As such, the prisoners’ ability to identify themselves as being a gang member was up to their interpretation whether it related to a more social definition or one of a legal nature (p. 802). Indeed, only 10% (n=39) of the prisoners noted agree or strongly agree with the item “ I consider myself to be part of a prison gang” (p. 807). Since the results were based on this single measure, the findings may be an indication of an inflation or underestimation of gang membership. In fact, the self-reports were not consistent with official documented figures of each establishment, which placed prison gang membership between 19% and 28% (p. 813). If the researchers had they used a better and accurate measure of gang membership, which included official prison data, or different representations of gang memberships, their results may have been a more accurate depiction of the realities. As a result possibly representing a stronger relationship between gang membership and other

variables.
In addition, the second threat to construct validity is the use of reactive measures, which can produce abnormal behaviour or responses throughout the study. Since the study took place during a lockdown not only did it affect who would be in their cells but it also affected the accuracy of responses (p. 807). Since a member of the staff administered the questionnaires, the prisoners may have been reluctant to answer the questions truthfully relating to perpetration and gang membership in fear of being disciplined. As such, we could presume that the prisoners may have changed their answers to one that is more socially desirable. Had a less reactive measure been used to identify gang membership or perpetration like using disciplinary infractions or official prison data, the findings may have been different. In other words, the different less reactive measures would have accurately captured the theoretical construct. As such, if the prisoners had answered truthfully, we would be more inclined to generalize the conclusions to other male prisons.
In sum, although the researchers’ findings suggest that there is a correlation between propensity to support prison gangs with aggression and other disruptive behaviors, gang membership and victimization, their findings are dependent on the measures they used. Moreover, had the researchers used less reactive measures as well used an adequate definition of the theoretical construct, we would be more confident that the results could be generalizable beyond the measures used. As such, the study findings are reflections of the measures used.