Interviews and qualitative analyses

By means of additional interviews it was sought to gain in-depth knowledge and better understanding of the patients’ experience of the intervention, and it was hoped that in addition to a deeper picture of patient motives, expectations and satisfaction, a better understanding of the process and outcomes of this procedure among this patient population could also be achieved. Semistructured interviews were utilized and qualitative content analysis was chosen as the method of analyzing the data in studies II and IV. The main aim in content analysis is to build a model to describe the phenomenon

in a conceptual form, and the analysis can be both deductive and inductive in nature (Elo

& Kyngäs, 2008).

In study II the follow-up data was obtained by semistructured individual thematic interviews containing questions about the participants’ recollections and experiences of the intervention group four years earlier. The participants were first asked to describe in their own words what they could recall from the group. Then complementary thematic interviews were conducted, where more specific questions about their experiences and memories of the group were asked. I conducted the interviews during the summer of 2005 and the interviews were 45-90 minutes long. All but one was audiotaped. Because one participant refused permission to audiotape the interview, his interview was conducted by making extensive notes. Deductive content analysis was used where the structure of the analysis was operationalized on the basis of previous knowledge (Elo &

Kyngäs, 2008); in this case using Antonovsky's Sense of Coherence Theory. As the SOC theory has been proposed for use in studying the process of psychoeducation (Landsverk

& Kane, 1998), deductive content analysis was considered appropriate way to test how the categories of SOC could explain the benefits that patients had experienced in the intervention. The interviews were first transcribed verbatim. The three major categories were then derived from Antonovsky’s Sense of Coherence Theory. The interviews were read and reread and the expressed benefits were extracted from the text and classified under the three main components of SOC. The categories captured the experiences of the participants very well, as they could be seen to represent a wide range of possible aspects of the intervention (cognitive, behavioral, emotional). Although in previous research comprehensibility has sometimes been seen as a cognitive, manageability as a behavioral, and meaningfulness as a spiritual and emotional component of SOC (see, e.g., Rabin, Matalon, Maoz & Shiber, 2005), some overlap between categories emerged. For example,

“gaining information” could be categorized as an experienced benefit in terms of both comprehensibility (more understanding about the illness) or manageability (information as a resource to cope better). Obtaining information or understanding was categorized under comprehensibility but if the answer referred more to new skills and behavioral aspects, for example, “learning to search for new information” it was categorized under manageability.

In study IV it was hoped that additional interviews would help to understand the motives for participating in the psychoeducation groups and give a deeper understanding of participants’ experiences in considering the group. I conducted the interviews in 2006 as an independent researcher. Because problems with auditotaping the interviews emerged in second phase of the study, all participant evaluations of the group experience were recorded through extensive notes during the interviews. During the interviews the participants were first asked to describe in their own words their motives for participating in the groups and then to describe their experiences. The open-ended questions from the evaluation questionnaire were also employed during the interviews (the best and worst things in the group experience, ideas of how to improve the group.

Inductive content analysis was utilized and the concepts were derived from the data.The answers from expectation and satisfaction questionnaires and interviews were combined and coded to cluster thematic aspects in each case and across cases. The transcripts were

then reread to confirm and refute evidence for each theme. The themes were then analyzed and sorted with regard to their content, and categories were created when they clearly emerged. Problems concerning the analysis of data were related to some vague, mainly psychotic answers, which were not possible to categorize under any theme or category. Citations from the interviews were used to increase the trustworthiness of the findings in both studies since the use of a co-reader was not permissible as the patients were promised that no one other than me would have access to the original interview data. (see Coffey, 2006; Elo & Kyngäs, 2008; Graneheim & Lundman, 2004.)

Table 3. Phases of the study

An exploratory RCT design and the study of patient motives,

expectations and satisfaction with the program

Year 2001 2005 2006

Aims To investigate the feasibility and effects of the pilot group psychoeducation intervention

To investigate the experienced long-term benefits of the pilot intervention from patient perspective

To study the effects of the intervention with an

Data collection Assessments with different outcome measures before and after the intervention

Semistructured thematic

interviews Assessments with different outcome measures at baseline, post-treatment and a 3-month follow up

Semistructured interviews, self-report questionnaires

Participants Forensic and challenging non-forensic patients with

Study design Quasi-experimental research design where the intervention group (n=7) was matched with a treatment as usual control group (n=8)

Qualitative interviews with patients who had attended the intervention group four years earlier

An exploratory RCT design to study the efficacy of the intervention

Qualitative interviews after the intervention and collection of self-report questionnaire data (n=34) before and after intervention

Deductive content analysis Statistical analyses:

- Fisher’s Z-test - Student’s t-tests - Mann-Whitney U test - Cohen’s d - Pearson’s correlation coefficient

Inductive content analysis Statistical analyses:

-Repeated measures ANOVA

Reporting Study I Study II Study III Study IV

6 Overview of the original

studies