ENHANCING STUDENTS’ PROBLEM -BASED LEARNING STRATEGIES THROUGH SELF-ASSESSMENT

Abstract

Active student participation is critical to successful Problem Based Learning (PBL). Self-assessment is one potential solution that might reduce students’ dysfunctional behaviour affecting the quality of PBL tutorials, which in turn impacts on the achievement of PBL goals. However, there is no convincing theory explaining how self-assessment works in PBL. Some self-assessment tools do exist; however, there is not yet sufficient validity evidence to support their underlying theoretical constructs. To fill this gap, this thesis explores the relevant theories and investigates the ways in which self-assessment can be promoted within the PBL process. This includes the development of a validated self-assessment tool that would assist students’ learning in PBL tutorials. The study was conducted using a mixed-method design in four stages among pre-clinical students at the Faculty of Medicine at the Islamic University of Indonesia (FM IUI), Yogyakarta, Indonesia. In the first stage, semi-structured interviews were conducted with 10 students and 10 tutors, who were all experienced in PBL tutorials, in order to investigate the relationship between students’ self-assessment and learning in the PBL process. The analysis was undertaken from a self-regulated learning (SRL) theory perspective. The findings revealed that students employed learning preparation by analysing the task and developing personal learning goals prior to the tutorial performance. During the tutorial performance, students displayed three skills: reasoning, teamwork and self-directed learning (SDL). Students conducted an unstructured self-assessment of their learning by applying self-observation, self-judgement and self-reaction. In conducting all of these activities, students required support from peers, tutors and the faculty. Students also found that PBL tutorials and self-assessment activities had impacts on their study behaviour. In the second and third stages, respectively, the development of the self-assessment tool was conducted via two processes: namely, Scale Construction and Scale Finalisation. First, 80 Likert scale items were developed based on the findings in stage 1 and consideration of scale items developed in previously published studies. Fifteen experts and 30 students were then invited to review the results. This process reduced the item pool to 37 items, which were then tested on 256 third- and fourth-year students. By applying item analysis and exploratory factor analysis (EFA), two factors were identified as underlying the developed tool. These factors theoretically connect with domains underlying the blueprints (reasoning, teamwork and SDL). As a result of this process, 29 items were retained. Next, 238 students from the first and second years used this 29-item tool. The data was then analysed using confirmatory factor analysis (CFA), by which a 2-factor model with 14 items was identified, with fit indices criteria showing a good fit of the model to the data. This model also showed good internal consistency (Cronbach alpha coefficient > 0.8) for each sub-scale (the Active Learning and Critical Thinking sub-scales). The scale, which is labelled as the Self-Assessment Scale of Active Learning and Critical Thinking (SSACT), was also stable when applied to two independent samples. The process of Scale Construction and Scale Finalisation provided evidence that the finalised tool (SSACT) had valid content and a good internal structure. Stage four was conducted to evaluate the initial learning impacts after a short implementation of the self-assessment tool (SSACT) through semi-structured interviews with 11 students. The findings indicated that the implementation of this tool in relation to the PBL tutorials helped students to apply self-regulatory processes. It also increased students’ knowledge about the tutorial process and increased students’ motivation in regard to improving their tutorial performance. The tool also empowered students’ positive behaviour during PBL tutorials. It enabled students to adjust their learning plans in terms of the learning strategy they applied during self-study. Interestingly, although the implementation of this tool was conducted as a research activity and without support from tutors or faculty, students became more aware of the kinds of support that they would need from tutors and faculty in order to maximise the benefits of the tool. Viewed through a framework that places SRL theory, expert learning and constructive-learning regulation concepts in a PBL context, the developed tool (SSACT) explicitly links the students' self-assessment strategies with their learning preparation and their tutorial performance, all of which constitute a learning cycle. Using the tool, students observed and judged their tutorial performance. They then took appropriate action in the next learning preparation. The learning cycle was mainly directed by students’ motivational and metacognition (knowledge and regulation) states. The tool appeared to stimulate students to regulate their learning by applying metacognitive regulation with the aim of improving their tutorial performance in order to meet the goals of the PBL. This was mediated by the increase in students’ motivation and metacognitive knowledge, and induced behavioural changes during the tutorial performance. The utility of the tool, in assisting students to improve their tutorial performance, required the availability of educational supports that further enhanced students’ motivation and metacognitive knowledge. The present study has resulted in a self-assessment tool with sufficient validity evidence. This work also provides an understanding of the connection between self-assessment and SRL. It extends the SRL model, expert learning and constructive-learning regulation concepts by placing these models in the context of self-assessment implementation in PBL tutorials. This work extends the SRL model by providing a greater understanding of the role of metacognitive knowledge and regulation in each phase of the model. It extends expert learning and constructive-learning regulation concepts in relation to the connection between motivation, metacognitive knowledge and regulation and their contributions in influencing students’ PBL tutorials.