RESUMO
This paper evaluates whether participation in the Thinking for a Change cognitive behavioral program produces improvement in social problemsolving skills in a prison context. Data are derived from a randomized experiment, with a focus on whether improvement in social problemsolving skills varies across modified delivery formats, and whether improvements are attributable to program completion or program dosage. We find that there are significant improvements in social problem solving between the pre- and post-test, and that delivery of the curriculum using video conferencing technology or inmate co-facilitated formats produces equivalent outcomes relative to traditional classroom administration. On average, significant improvements accrue to participants who receive greater program dosage. However, program completion, commonly viewed as a primary marker of satisfactory program performance, is not associated with improvement in social problem solving. Implications of the findings for rehabilitation programming are discussed, including considerations in a pandemic context.