3.6. Procedural Stages in the Present Q Methodological Study
3.6.3. Q Set Design and Content
3.6.3.1. Sampling methods in the current study
A hybrid-sampling technique was employed in the current study to generate items for the Q Set. This involved incorporating aspects of naturalistic and ready-made samples. The naturalistic items were derived from questionnaire data and the ‘ready-made’ items were derived from other discourse surrounding the topic of ‘reintegration’.
3.6.3.2. The generation of items
The primary method of finding items for the Q set was through analysing questionnaire data. I considered this sampling technique to be the most important as it incorporated the views of CYP themselves, which was consistent with the emancipatory nature of the research. It was also deemed important to canvass the perspectives of the key systems surrounding the CYP (family and school), which fit coherently with the bioecological theoretical framework (Bronfenbrenner, 1979) underpinning the research.
Questionnaires were distributed to five participants, their parent/carer, and a member of school staff who knew the young person well and who had been involved to some extent with their reintegration. Three questionnaires were therefore returned in relation to each participant. The questionnaires were
qualitative in nature and used one open-ended question. Specifically, participants, parents and staff were asked to describe three factors they believe had supported the young person’s successful reintegration into mainstream education. (The questionnaire data is displayed in Appendix H and the three questionnaire templates are contained in Appendix I).
According to Watts and Stenner (2012), one of the main criteria for an effective Q set is the coverage of the topic area. If this criterion is not met successfully, participants’ viewpoints can be restricted and certain aspects of their viewpoints may not be able to be expressed as a result of statements not being included. To ensure that the Q set was suitably balanced and to improve coverage, I compared the questionnaire data with a literature review until a ‘saturation point’ was established where the population of potential statements had been fully sampled. Using McKeown and Thomas’ (1988) terminology, this resulted in the ‘ready-made’ aspect of my hybrid sample. In addition, I consulted professionals within the field (colleagues in the EPS and staff in APs/mainstream schools) in an attempt to gain professional views and perspectives.
A ‘structured Q set’ was created, which involved employing a technique to identify themes emerging from the questionnaire data and literature. This ‘thematic analysis’ involved colour coding the data obtained from the questionnaires and literature in order to identify overarching themes. From the questionnaire data, ten themes were identified in relation to the factors that pupils, parents and school staff believed supported the pupils’ successful reintegration. Ten themes were also identified from the literature review, however there was significant overlap with
those derived from questionnaire data and therefore similar themes were grouped together. Items (statements) were then generated relative to each theme and the final Q set contained a representative number of items covering aspects of each demarcated area. This was deemed advantageous in ensuring that a clear sense of system and rigidity was brought to the sampling process, resulting in a balanced, representative and non-biased Q set. At this stage of process, the Q set consisted of 90 items. (Appendix J and K display the colour-coded literature review and questionnaire data respectively. Appendix L contains a table of the themes and statements derived from each source).
My original intention was to use direct quotes from the questionnaire data and literature to form statements for the Q set. This would have followed the advice of Brown (1980), who advocated changing the source of each item as little as possible. However, this proved to be unattainable as many of the responses contained within the questionnaires were personal and specific, and the language contained within the academic literature was often technical or deemed inaccessible to children. It was also important to be mindful of attaining coverage of the topic area while limiting the statements to a suitable number. Furthermore, I followed the advice of Watts and Stenner (2012) for generating suitable items for the Q set. These authors suggest that each item should contain a single idea, avoid negative and complex terminology, and begin with the same prefix where possible. Taking direct quotes from questionnaires and academic literature would have deviated from these guidelines. The items in the Q set were therefore based as much as possible on statements from the concourse, however all had to be adapted.
3.6.3.3. Refining the Q set
Once a set of 90 items had been identified from the concourse, duplicate items were removed and those that were similar were combined and revised. This resulted in a Q set comprising 50 items. Watts and Stenner (2012) note that the exact size of the final Q set is largely dictated by the subject matter, however a Q set of somewhere between 40 and 80 items is typical. There is debate in this area, with some authors suggesting a wider range of between 10 and 100 items (Cross, 2005) and others demonstrating that as few as 25 are satisfactory (Watts & Stenner, 2005). Indeed, Watts and Stenner (2012) argue that in some circumstances it is sensible to employ a more limited number of items, for example when there is a requirement to make the sorting task less taxing. This informed my decision to reduce and refine the Q set further to ensure that it was not too demanding for CYP.
In order to refine the sample of 50 items, a member of staff at the Tavistock and Portman NHS Trust (who was familiar with Q methodology) and a colleague within the EPS (who had experience working with excluded pupils as well as conducting psychological research) were consulted, and each provided comments and suggested refinements to the Q set. From these consultations, items that were deemed to be too ambiguous, poorly worded or incomprehensible to children were removed or adapted and all statements were standardised so that they began with the pre-fix “it helped me when…” Following refinement, the Q set consisted of 36 statements, which was deemed to be more manageable for the current participant sample, whilst also maintaining a broad coverage of the topic area.
The final stage of finalising the Q set involved conducting a pilot study with a participant in a mainstream secondary school who had experience of transition. The aim was to trial the statements; obtain feedback regarding the number of items and terminology used; and to identify any potential issues in terms of practicalities when completing the Q sort. The pilot participant reported that the number of items was adequate and that the wording of the statements was comprehensible. In addition, it was reported that the activity of Q sorting was enjoyable and a “cool way to say
what you think” about an issue. From the pilot study, one item was added to the Q
set, which the participant felt was particularly important in supporting a successful reintegration into mainstream education. This particular statement had previously been removed due to its potential ambiguity, however I felt that it was essential to listen to the view of an expert in the field (the participant) and the statement was subsequently re-included. This resulted in a final Q set comprising 37 statements (see Appendix M).