MAISHA Cluster Randomised Trial - Primary Analysis Dataset

The MAISHA study (component A) is a mixed methods evaluation, including a cluster randomised trial, of a small group based social empowerment intervention to prevent intimate partner violence (IPV) against women in Mwanza city, Tanzania. The trial was conducted among women already participating in a group-based microfinance scheme, with the microfinance group being the unit of randomisation and (in the intervention arm) the unit of intervention delivery. The CRT involved two surveys of trial participants, the first conducted prior to randomisation (baseline), and the second 29 months later (follow-up). The surveys collected data on demographics, childhood experiences of abuse, health, attitudes and norms relating to gender and IPV, relationships, IPV experiences and responses, children, and respondents' interactions with their communities. This dataset underlies the primary trial analysis of intervention impact. It comprises 1,049 records (each pertaining to an individual trial participant), and contains a subset of variables from the baseline and follow-up surveys. NB: Data pertaining to a second trial of the same MAISHA intervention conducted among women not receiving microfinance (MAISHA component B) are not included in this record. The questionnaires (baseline and follow-up) were developed in English, translated into Swahili, and independently back-translated into English for validation. The violence questions were adapted from the WHO Violence Against Women instrument, which has also been widely used in demographic and health surveys and other prevention intervention trials of intimate partner violence. Interviews were done face-to-face, in private, by female interviewers trained in interviewing techniques, gender issues, violence and ethical issues related to research on intimate partner violence. Responses were entered directly onto electronic tablets.