Establishment in productive occupations and perceived work ability among former students with special educational needs one year after upper secondary education

2.1 Design

This descriptive study, which investigates productive occupations and perceived work ability among former students with SEN, was conducted using an embedded mixed methods approach [24]. Quantitative cross-sectional data from a questionnaire was used to investigate productive occupations among the former students, supplemented with qualitative data collected using the Swedish version of the Worker Role Interview (WRI) [25] to describe their perceived work ability. This study has been conducted in accord with the Declaration of Helsinki of 1964 and its later amendments. The Regional Ethics Board in in Linköping, Sweden, gave ethical approval for the study, study codes 2013/409-31 and 2015/203-32.

2.2 Participants and data collection

The study population was derived from an earlier study by Yngve et al. [21], evaluating the impact of a student-centred ICT intervention on school participation among 300 upper secondary school students with SEN in Sweden. These 300 students were from all three upper secondary educational levels and attended either a vocational or higher education preparatory programme. Of these 300 students, 244 provided written informed consent and contact information and agreed to be contacted one year after leaving upper secondary school, making them eligible for inclusion in the present study. The data collection was performed during year 2014 until 2018 and consisted of two parts. First, eligible participants were sent the cross-sectional questionnaire, in which they decided whether or not to participate in the WRI interview, which was conducted on a later occasion.

Contact information for 22 students was invalid at the time of data collection, resulting in 222 eligible participants, who were contacted by telephone and/or via email approximately one year after they left upper secondary education. Seventeen students declined participation when approached. The questionnaire was answered by 81 students while 124 did not respond, resulting in a response rate of 36%. No significant differences in demographics (age, diagnosis, educational programme, gender or native language) were found among the 81 participants and the 141 students who did not respond to the questionnaire. Forty-one (51%), of the 81 participants, gave consent to participate in the WRI interview, of whom 20 actually participated. No significant differences in demographics were found among the former students who participated in the WRI interview and the 141 students who did not respond to the questionnaire.

2.3 Data analysis

Quantitative, cross-sectional data consisting of responses to the questions from the questionnaire and ratings of WRI items was analysed using descriptive statistics [30]. In the present study, productive occupations are defined as including paid and unpaid activities [26] which in turn include paid employment, studying, and activities to improve opportunities for employment, such as labour market training and work experience placement. Participants were divided into two groups based on the type and extent of their reported productive occupation. Participants who reported that they on a weekly basis worked 50% or more of full time, were enrolled in higher education or who currently were on parental leave from a specific workplace (n = 1) were considered to be ‘Established in work or higher education’. ‘Not established in work or higher education’ included participants who were working less than part time (50%) on a weekly basis, were unemployed/seeking employment, enrolled in adult education equivalent to upper secondary education, or were receiving compensation from the Swedish social insurance system. Fisher’s exact tests [30] was conducted to investigate the differences between the two groups in terms of demographics, final grades and type of ‘time-assisting’ or ‘supporting writing and reading’ ICT devices received during the intervention in upper secondary education. Analyses were performed in SPSS, version 24 [31], and the significance level was set at p < .05.

The written notes on the WRI rating forms constituted the unit of analysis in the qualitative content analysis. Since the WRI is theoretically based on concepts in the MOHO, a deductive approach was followed, guided by Graneheim, Lindgren and Lundman [32]. A deductive analysis moves from theory to the concrete data level [32] and concepts from the MOHO were used to guide the analysis of which aspects influenced participants’ perceived work ability. The six MOHO concepts in the WRI were used as content areas and were, together with their theoretical definitions, inserted into matrices of the related theoretical constructs of motivation, daily routines and social environment, which were used as categories. For each content area, the manifest content of the written notes related to the included WRI items was assembled. As an example, the manifest content from the items ‘Assesses abilities and limitations’, ‘Expectations of job success’ and ‘Takes responsibility’ was assembled in the content area ‘Personal causation’ located under the category ‘motivation’. The content of each area was read through several times by the first author to obtain a sense of the whole. The process of abstraction began with identifying meaning units in the written notes. In most cases, the written notes were abstracted to even shorter, concrete meaning units, which were then coded. Meaning units that concerned the same thing, irrespective of whether they were considered as supportive or hindering factors for the person's work ability, received the same code, inspired by the theoretical concepts. The analysis and process of abstraction were performed with a word processor program by using matrices and highlighting the text in different colours. In the process of abstracting codes in relation to the concepts, the first and last author engaged in dialogue to reach consensus.

Most participants had become ‘Established in work or higher education’ (n = 51, 63%), with 43 participants working, seven enrolled in higher education and one being on parental leave from a specific workplace (see Table 1). In the group ‘Not established in work or higher education’ (n = 30, 37%), the largest proportion of participants were seeking employment (n = 12, 15%) or enrolled in adult education equivalent to upper secondary education (n = 10, 12%). On a weekly basis, 64% of the participants (n = 37) reported full-time engagement in their current occupation, see Table 1.

3.2 Differences between participants ‘established in work or enrolled in higher education’ and ‘not established in work or enrolled in higher education’

Participants who were established in work or enrolled in higher education reported final grades in all subjects to a higher extent (p = .02 Fisher’s exact test), and had less often received time-assisting ICT (p = .01 Fisher’s exact test) than those who were not established in work or higher education. No differences between the groups were found in relation to demographics, or received ICT for writing and reading, see Table 1.

3.3 Participants’ perceived work ability

Eleven of the 20 participants who participated in the WRI interview were boys and 13 of them had become ‘Established in work or higher education’. The ratings of the participants’ WRI items are presented in Table 2. The WRI item ‘Commitment to work’ was the only item rated as supportive among all participants, indicating that work was a valued occupation for all. The item ‘Pursues interests’ was the item rated as most unsupportive in relation to participants’ psychosocial work ability, as almost more than a third of the participants had that item rated as interfering, which indicates that these participants experienced difficulties in finding ways to be satisfied with their daily lives, see Table 2. Items related to participants’ ‘Roles’ in and outside their productive occupation were rated as supportive for work ability for all but one participant. In regard to habits, also related to the content area ‘Daily routines’, hindering aspects in ‘Daily routines’ that interfered with perceived work ability were found by four of the participants. Items within the content area ‘environment’ had drop-out in several items because some participants did not have a specific workplace to relate to, but the item was for the most part found to be supportive of participants’ work ability. The WRI item that interfered most was related to participants’ perceptions of their manager, which was rated as interfering by three participants.

The six theoretical constructs included in the WRI that constituted the content areas were sorted under the three categories of motivation, daily routines and social environment, derived from the MOHO. The associated codes for each content area are listed in Table 2 and written in italics in the text under the following headings.

4.1 Methodological considerations

Some methodological considerations need to be addressed concerning the drop-out rate and trustworthiness of the qualitative findings. A longitudinal approach is often associated with an increased risk of a high drop-out rate and this study had a drop-out rate over 60%. Part of the explanation for this might be potential participants’ unwillingness to answer phone calls from unknown telephone numbers, or that their contact information had been collected several years previously. Measures were taken to combat this issue by using both phone calls and a web-based version of the questionnaire, with reminders. However, if participants had changed their telephone number or email address, these measures were inadequate. Access to potential participants’ personal identity numbers would have provided better opportunities to make contact with individuals who had consented to participate in the present study.

Furthermore, the small number of former students who participated in the WRI interview made it impossible to statistically compare if there were differences in the WRI ratings between those who were established or not in productive occupations. The credibility of the qualitative findings might have been strengthened because of the heterogeneity in the data, including WRI interviews with participants of both genders with experiences of SEN during upper secondary education, and who were and were not established in work or higher education. Credibility is also about the data collected and the appropriateness of the collection method [40]. Using the WRI, which was designed to investigate perceived work ability, provided a structured method to collect relevant information to investigate participants’ perceived work ability in terms of motivation, daily routines and social environment. Furthermore, using the MOHO to guide the content analysis provided a valid frame of reference and was considered helpful since it offered a theoretical explanation for how motivation, daily routines and the surrounding environment influence an individual’s perceived work ability [26]. However, an already processed text, such as the notes in the WRI rating forms, limits the possibility of abstraction and interpretation [41] and the analysis and presentation of the results was therefore kept at a descriptive level, which remained close to the participants’ statements. As pointed out by Hsieh and Shannon [42], a deductive approach might lead to new insights being missed. Therefore the first and last authors engaged in an ongoing dialogue during the process of analysis and formulation of findings, to ensure correspondence between the data and the theoretical model. Another concern, related to dependability [32], is the fact that there were three different WRI interviewers. However, both the first author and the research assistants had knowledge of the MOHO and were trained users of the WRI with its semi-structured nature, which might have strengthened the dependability of the collected data.