Primary outcome: Self-determination
Self-determination was the primary outcome of the study because of the previously established correlation between high levels of self-determination and post-school employment and education [
50‐
52]. There was no change in the total self-determination score found in this study. A potential reason for this may have been the varied levels of adherence to the BOOST-A™, as the average number of modules completed was three indicating many participants did not complete the My Progress module. Another explanation may be the lack of face-to-face training in how to use the BOOST-A™, which was delivered remotely via an online platform. A meta-analysis of the effectiveness of technology-based programs for adolescents on the autism spectrum found that programs that were entirely self-directed by participants had a smaller effect than programs administered by a specialist [
32]. Whilst there is a need for programs that are not only effective but also easily accessible, affordable, and user-friendly [
32], technology should not be used as a substitute for face-to-face support [
53]. Therefore, ensuring direct access to a trained professional to facilitate use of the BOOST-A™ may be an important consideration for future iterations of the program.
A significant difference was found in self-determination between the intervention and control groups was in the Home subscale. This may suggest that the BOOST-A™ supported parents to provide increased opportunities for the adolescent to practice decision-making, goal setting, and problem solving in the home environment. This finding is of interest, given that current literature tends to focus on school as the context to improve adolescents’ self-determination skills, with less focus on the home environment [
54]. In addition, the majority of existing transition planning programs focus on supporting school staff to enhance the self-determination of students with disabilities in the school environment [
13,
17,
55,
56]. However, parents are possibly the most consistent and enduring influence in their adolescent’s life, especially during the transition from school into post-secondary education or employment [
54,
57]. Parents model self-determined behavior in the home environment and provide opportunities for adolescents with autism to make choices; take appropriate risks; and develop skills in problem solving, self-regulation, and assertive communication [
58]. A strength of the BOOST-A™ is that it can be used either at school or at home and can be championed by parents and/or teachers.
The increase in opportunities provided at home reflects a potential shift in parents’ expectations for their children, as supported by the results from the process evaluation of the BOOST-A
TM [
59]. Parents who hold high expectations for their adolescents with autism can increase the adolescent’s self-determined behavior and improve their post-school outcomes [
15]. Furthermore, increased frequency of discussions about post-school plans in the home environment has been correlated with increased participation of adolescents on the autism spectrum in transition planning meetings at school [
60]. Therefore, changes in the behavior of the parents may result in increased opportunity to engage in transition planning for adolescents on the autism spectrum.
A possible explanation for the observed increase in the Home subscale of self-determination but not the School subscale is that parents were the primary contacts in this trial and the key point of liaison with the research team. Further research in this area might assist in understanding the relationship between the home and school settings, and the opportunities for self-determined behavior provided to adolescents on the autism spectrum in these settings.
Secondary outcomes
The BOOST-A™ led to a significant increase in career awareness among the adolescent participants. Career awareness is defined as the level of engagement with external sources of career information, such as parents, teachers, and written information, as well as the adolescents attitude towards these sources of information [
40]. Career awareness is predictive of being productively engaged in education and employment in the first year out of school [
61,
62]. The finding that the BOOST-A™ increased career awareness supports the hypothesis that adolescents on the autism spectrum who use the program may have an increased likelihood of transitioning to post-secondary study and employment after school. No significant differences were found in career planning, which is the amount of planning that has been completed [
40]. The lack of significant increases in career planning may have been because not all participants completed the fourth module that supported them to revise goals and progress through planning.
There was also a significant increase in transition-specific self-determination favoring the intervention group. The Transition-specific Self-determination scale looked at the adolescents’ opportunity for active engagement in transition-specific team meetings; exploration of interests and strengths; goal setting; and real-life experiences, such as work experience, mentoring, and part-time work. However, the psychometric properties of the Transition-specific Self-determination scale are currently unknown and so these results should be interpreted with caution. Future research to validate this scale is recommended.
For both groups, adolescent quality of life decreased, whilst happiness with life as a whole increased. The overall reduction in quality of life during adolescence is consistent with a decrease in quality of life that is seen in mid to late adolescence for the general population [
63]. This decrease in quality of life is likely because adolescence is a period in which young adults experience a shift in roles and seek greater independence, which is often at odds with their continued dependence on caregivers [
63]. In addition, adolescents are presented with many new challenges as they transition out of high school that are likely to impact on quality of life. Overall quality of life for both groups was below the normative range for Australia, which is between 73.4 and 76.4 points out of 100 [
42]. Evidence suggests that quality of life among people on the autism spectrum is lower than that of people without autism across the lifespan [
7]. Therefore, further research is warranted that looks at quality of life during the transition period for adolescents on the autism spectrum, and how this compares to adolescents without autism.
There appeared to be a discrepancy in that overall quality of life decreased whilst happiness with life as a whole increased. One possible explanation is provided by the concept of the ‘just right challenge’ in self-determined learning theory, which describes how opportunities should provide an optimal level of challenge to enhance adolescents' capacity to regulate their feelings and actions [
64]. Whilst encountering challenges in the transition planning period, adolescents may describe a reduction in overall their quality of life but an increase in happiness as they learn new skills and overcome challenges. Parents in this study rated their adolescent’s quality of life lower than the adolescents’ self-ratings. This finding is consistent with previous research that proposed adolescents on the autism spectrum may perceive the difficulties they face to be less of a problem than their parents [
7]. In addition, parents may have made assumptions on the meaning of a good life without asking their children what would make them happy, which is an important prerequisite for emotional wellbeing [
65]. Discrepancies between parent and adolescent perspectives of quality of life is an issue requiring further exploration in future research.
Overall, only one area showed a significant difference as reported by the adolescents, in comparison to three areas as reported by parents. The lack of significant differences as reported by adolescents is noteworthy, because the BOOST-A
TM aimed to improve adolescents’ perceived autonomy and control, as this has been linked to improved post-school outcomes [
66], and increased subjective quality of life [
63]. This indicates more work may need to be done to improve adolescent outcomes in the transition planning process using the BOOST-A™.
The lack of between-group differences for many of the outcomes suggests that the BOOST-A™ was more helpful for some adolescents on the autism spectrum than it was for others. This finding may be due to the range of characteristics of people on the autism spectrum and is consistent with evidence that there is a wide variability in outcomes for children on the autism spectrum [
67,
68]. Another potential reason for the varied results for participants could be that some had comorbid diagnoses such as attention deficit hyperactivity disorder and anxiety, which could have introduced additional considerations related to transition planning. A process evaluation was performed immediately following the quasi-randomized controlled trial to determine individual characteristics and contextual factors that support positive outcomes from using the BOOST-A™. The process evaluation found that whilst the BOOST-A™ supported some adolescents to engage in the transition planning process and develop new insights that led to clearer plans for the future, barriers included not having access to a professional to guide the way and difficulty motivating the adolescent to engage in the process [
59]. Full results of the evaluation are reported separately [
59].
Limitations
A limitation of this study was that the participating adolescents’ autism diagnosis was based on parent-report and confirmed by the SRS-2 [
35]. Ideally, the Autism Diagnostic Observation Scale (ADOS) [
69] would have been used to verify autism diagnosis, since it has good sensitivity and specificity [
70]. However, this was not possible because the ADOS is administered face-to-face and study participants came from a wide spread of geographic locations across Australia. The study had a low attrition rate for parents (9% control; 12% intervention), but higher for the adolescents (10% control; 31% intervention). Consequently, the final analysis was based on a sample containing less than 40 adolescents in each group, which may have resulted in the study being underpowered to detect between-group differences for all the outcomes measured.
Use of a quasi-randomized and non-blinded treatment allocation presented potential sources of bias. However, baseline comparisons revealed no significant differences between the control and intervention groups in outcomes and demographic variables other than socio-economic status. A general linear model confirmed that the significant between-group differences for the intervention effects remained after adjustment for socio-economic status. Given the difference in socio-economic status, it would have been optimal to have collected information about parents’ academic qualifications and professions to determine if there between-group differences in these areas. Although participants were excluded if they were participating in any other formalized transition planning program, it would have been optimal to gather data from the control group about any informal transition planning that may have been initiated by parents or schools during the study period. Different types of regular practice undertaken by the control group may have influenced their outcomes, so this study may have underestimated the true impact of BOOST-A™ over a standardized control group with only basic transition planning.
The timeframes for this study did not allow for follow-up to determine whether the effects of the intervention were maintained or to gather information about participants’ employment outcomes after graduation from school. This is a limitation because career readiness outcomes, such as self-determination, that were used in the study are only correlated with employment. Their observed improvement in this study may not necessarily lead to an increase in employment [
71]. The use of employment as an outcome would have reduced the risk of bias introduced by the use of self-report measures, such as social desirability. It is recommended that future studies are of sufficient duration to explore the maintenance of the changes in career exploration and self-determination over time, as well as post-school employment outcomes after using the BOOST-A™.
Strengths of the current study included the use of multiple raters (parents and adolescents), blinding of participants to trial hypothesis, as well as self-report measures that eliminated the need for blinding of evaluators. The inclusion of a control group in the study ensured maturation bias did not influence results, especially given the 12-month duration of the study.