2.4. Analysis

Giorgi’s method of analysis [50], which involves creating a series of categories and subcategories, was used to analyze the information from both the in-depth interviews and the focus groups. This procedure was carried out in several phases. The first one was an in-depth reading of all the discourses, which had already been transcribed verbatim. The second phase involved a second reading and the division of the data into parts. The basis of the division into parts is meaning discrimination, which presupposes the prior assumption of a disciplinary perspective (social work, in this case). These meaning discriminations constitute parts known as meaning units. The meaning units were examined, tested, and redefined so that the disciplinary value of each unit could be more explicit. These meaning units were then grouped into broader categories according to their shared characteristics and the disciplinary value. In the last phase, the contents of each of the categories were interpreted and analyzed based on the phenomenon or experience lived.

The theoretical–methodological approach was adequate to achieve the objectives of the study. The data obtained were relevant in the context and in other contexts, when compared to the literature.

As for the validity of the results of the analysis, contrast through triangulation was used to control for potential biases resulting from the heterogeneity of the data and the informants’ different points of view. To make a contrast between the differences and similarities conveyed in the discourses, the techniques of focus groups and in-depth interviews were used. In terms of triangulation between subjects, informants were selected from different settings and fields of work to diversify the information present in the discourses regarding the participants’ work experiences in these services. Two researchers began the analysis after the first interview in order to constantly verify that it was in line with the study’s objectives and in order to be prepared in case any change in the research design was needed (it was not). The main categories that researchers identified in the analysis were shared with the participants (by email) to confirm the discourses. In the participants’ discourses where contradictory information was detected, this moment was used to clarify it. The analysis was shared with the rest of the team to ratify the categories. At the same time, an external researcher (with expertise in the subject) validated the analysis.

Reflexivity and a self-critical attitude were maintained throughout the process by all the researchers. To avoid influencing data collection, sample recruitment, and location, the researchers only knew the topic in a superficial manner (as health professionals) and it was not their usual work/subject matter of research.

Atlas.ti 8.0 software (Scientific Software Development GmbH, Berlin, Germany) was used to analyze the discourses.