*4.3. Procedures*

To begin, we requested formal authorization from the VSO to contact users of the institution who expressed interest in participating in the research. In parallel, the research project was submitted to the University's Ethics Committee for advice and recommendations.

After collecting all authorizations, we began the study following the recommended ethical and deontological principles. All data and information collected during the investigation were stored on a computer within the VSO, accessible only by a code provided to the researchers. The interviews were conducted in the presence of a VSO psychologist during their planned appointments, with data collection limited to the selected instruments. The participants were informed about the purpose of the study, its procedures and also relevant issues of anonymity and data confidentiality; following this, they signed the informed consent. Interviews lasted approximately 30 min and were audio-recorded.

The data treatment and analysis procedure was predominantly qualitative. Data from the interviews were fully transcribed and then subject to qualitative analysis, giving rise to the categories and subcategories of analysis. To this end, we used an interpretive and inductive method of data treatment, based on the strategies of grounded theory (Charmaz and Henwood 2017), according to which the categories emerge from the subjects' narratives, their experiences, and the meanings constructed. Thus, taking a "sentence" as a unit of analysis, we proceeded to categorize it according to the semantic records. To ensure the reliability of the results, steps were taken: (i) saturated coding of all the data collected through the analysis was carried out independently by two researchers (the field researcher and an experienced supervisor); (ii) joint dual analysis was performed, based on designations emerging from the participants' narratives, via a continuous debate to scrutinize categories; (iii) inclusion of a third researcher to establish an agreemen<sup>t</sup> between the two coders in respect of the final designation of emerging categories and subcategories.

Table 2 shows the categories and subcategories that resulted from the qualitative analysis, expressing the reasons why mothers who were victims of violence remained in the relationship and the reasons that make them leave the relationship and ask for help.


**Table 2.** Sociodemographic characterization of the sample (N = 15).
