Method: A mixed-method study design, sequential exploratory approach was chosen to develop the instrument. Initial semi-structured interviews were conducted with 4 preceptors at the teaching hospitals. Five main themes emerged from the interviews. The themes were developed into a 23-item questionnaire. Nineteen Head or Assistant Head of clinical departments were asked to review the relevance of the content. The questionnaire was later sent out to 34 preceptors and 35 paramedic staffs to participate in the construct validity study by conducting exploratory factor analysis (EFA). SPSS version 21 software was used to analyze the data and Varimax rotation method was performed to simplify and describe the data structure.
Result: Review of the factor structures suggested that the most appropriate fit was 5 factors. Most of the questionnaire items were relevant to assess performance (4.65 + 0.15), except in item 4 of the clinical skill factor. The 23 items of the evaluation instruments showed that five factors were extracted which explained 73.9% of the variance between them. Construct validity was achieved after the instrument was run for an iteration of eight times, with Cronbach's alpha of 0.951.
Discussion: The instrument has achieved the desired content and construct validity score. It can be used by other institutions to assess their medical graduates' performance in the clinical setting.
METHODS: This study employed mixed method in a sequential exploratory design. The semi-structured interviews were conducted with three graduates from different cohorts. The qualitative analysis of the interviews found six emerging themes for professional behavior and clinical competencies development. These themes were then developed into a 55-item questionnaire. The questionnaire was then distributed to 84 medical graduates for exploratory factor analysis (EFA) from February to April 2019. The quantitative data were analyzed using IBM SPSS ver. 21.0 (IBM Corp., Armonk, USA) for principal axis factoring. After conducting EFA, we proceeded with confirmatory factor analysis (CFA) with another 120 graduates to validate the tool.
RESULTS: Eighty-four graduates completed the questionnaire for EFA. Upon completion of EFA, 35 out of 55 items of the questionnaire were found to be valid and reliable. The most appropriate fit was seven factors, which explained 58.18% of variance between them after 15 iterations with Cronbach's α of 0.916. The personal satisfaction factor was noted to be weak. It was therefore added to patient management factor due to its similar intention. The final EFA factor after the modification was six. The CFA found that 34 out of 35 items was valid and reliable that representation of the latent variables.
CONCLUSION: The questionnaire has achieved the desired construct validity score and can be used as an evaluation tool to assess professional behavior and clinical competencies from the graduates' perspective.