Revisiting Collective Efficacy: Insights from Cognitive Interviews and Experimental Surveys

This three-part study revisits the collective efficacy scale developed by Sampson and colleagues in the 1990s to see if the scale needs updating. Defined as community cohesion and social control, collective efficacy is often used in community research to explain crime rates. The first phase uses cognitive interviews to understand how survey respondents interpret the original collective efficacy survey. The interview data then helps create new versions of the ten questions in the second phase. In the third phase, ChatGPT generates a revised version of the collective efficacy scale. The study employs a split-ballot experimental survey to test whether the revised versions improve reliability and validity. Each of the 258 respondents is randomly assigned one version of the scale. Various data quality metrics—such as internal consistency, nonresponse, response time, straightlining, and concurrent validity—are used to identify the best survey version. While the results do not conclude that one version is superior, they do favor the revised ones, indicating that there is room for improvement over the original scale.