Qualitative assessment of medical students’ perspectives toward PWD
Concept elicitation data suggested seven themes directly linked to medical students’ perspectives toward PWD. These themes were multidimensional, intersecting, and overlapping key topics included: Comfort interacting with people with disability in society; Negative impression of disabled for themselves; Conditional positive attitude; Impression of social attitude towards disabilities; Need for education; Support of disabled by health professionals and institutions; and Comfort working with people with disabilities in a clinical setting. Overall, there were no large differences in concepts between pre-clinical years students (Med 1 and 2) versus clinical years students (Med 3 and 4), except for Comfort working with people with disabilities in a clinical setting, that was addressed only by students with clinical experience.
Themes that were mentioned by the participants in the qualitative study illustrated the themes obtained by the validation of the MSA-PWD (see Appendix 1 for more details on themes, and quotes from participants). Thus, the content of the MSA-PWD scale was validated in the local setting.
Quantitative assessment of medical students’ perspectives toward PWD
Sample characteristics
This study sample was randomly drawn from a total population of medical students of 251 individual. The number of students is evenly distributed across enrollment years, in addition to an even gender distribution in each enrollment year. In this sample of 143 medical students, the majority were of female gender, reported having structured experience working with PWD and did not have a relative with disabilities. The mean age was 22.6 (1.42) years. They were distributed across enrollment years as follows: 33 (23.4%) students in med 1, 58 (41.1%) in med 2, 20 (14.2%) in med 3 and 30 (21.3%) in med 4 (Appendix 2).
Factor analysis
Construct validity and reliability of “Medical Students’ Attitudes Towards Patients with Disabilities” scale (MSA-PWD)
Firstly, all the items used by Symons et al. were included in the factor analysis. This process led to a factor structure that is not coherent or comprehensible and could not explain the objective of the scale construction (details in Appendix 4). Then the items included were items 3 to 20 from the original tool were included. The clinical scenarios A and B were excluded and used as separate questions because their use was not convenient in the same factor analysis (inadequate loading on respective factors) (Appendix 4).
Based on the above, the adjusted MSA-PWD’ rest of the items were included again in the factor analysis. The total scale, which ranged from 33 to 61, was calculated by summing all the answers, with higher scores indicating a more positive attitude towards PWD. The factor analysis using the rotated component matrix technique was used to test the construct validity of MSA-PWD and ensure the model’s adequacy. All items of the scale could be extracted from the list, except for items 8 and 11 that were also removed from the final scale [because of over-correlation (r > 0.9), had a low loading on factors (< 0.4), or a low communality (< 0.4)].
The final scale comprised sixteen items; it measured different facets of medical students’ attitudes towards disabled patients. The Kaiser-Meyer-Olkin measure of sampling adequacy of the final scale was 0.827, and Bartlett’s test of sphericity was significant (p < 0.001). Its total Cronbach alpha was 0.816 (Table 1). The final scale yielded four factors, with the first factor explaining 35.76%, the second explained 13.51%, the third 8.91% and the fourth 7.29% of the total variance (65.50%). Factor 1 includes “comfort interacting with people with disability in society”; factor 2 involves “negative impression of disabled for themselves”; factor 3 is about “conditional positive attitude”; factor 4 explains “impression of social attitude towards disabilities” (Table 1).
The four factors yielded by the factor analysis converged adequately with the first four themes produced by the qualitative arm of the study. Furthermore, these results confirmed the contents of the scale constructed by Symons et al. although yielding a different factor structure,
Validity and reliability of the confidence dealing with PWD and ableism self-assessment index
A factor analysis using the rotated component matrix technique was used to test the construct validity of the confidence dealing with PWD self-assessment tool and ensure the model’s adequacy. All items of the scale could be extracted from the list, and no over-correlation (r > 0.9), or low loading on factors (< 0.4), or a low communality (< 0.4) were detected. The Kaiser-Meyer-Olkin measure of sampling adequacy was 0.659, and Bartlett’s test of sphericity was significant (p < 0.001). The total Cronbach alpha was 0.811 (Table 2). The index yielded one factor, explaining (64.44%) of the total variance.
Reliability of the clinical experience of the students scores
The clinical experience and ease of dealing with different clinical situations constituted a theme in the qualitative data results and was assessed in the quantitative analysis via 2 scores. “Scenario A” for dealing with non-PWD, with Cronbach alpha 0.858 and “Scenario B” for dealing with PWD, with Cronbach alpha 0.539. (Appendix 3) Although these two scenarios were key aspects in the qualitative assessment, they could not be integrated in the MSA-PWD factor analysis. Therefore, they were useful as separate tools in convergent validity testing.
Convergent validity of the MSA-PWD scale
The MSA-PWD total scale and factors were tested for convergent validity with the confidence dealing with PWD self-assessment index and clinical experience scores. Pearson’s “r” demonstrated a positive correlation for the total MSA-PWD and its factor 1 (Table 3).
The different factors showed various levels of correlations and statistical significance, depending on the setting similarities of dealing with PWD and the students’ attitudes and levels of confidence dealing with PWD. Factors 2, 3 and 4 showed lower significance levels.
Known groups validity of the MSA-PWD scale
The MSA-PWD factor 1 could differentiate between different medical students’ classes, for which statistically significant results were reached. The trend test showed that MSA-PWD “factor 1” mean increased with higher enrollment year (Table 4). Similar results were obtained for confidence dealing with PWD self-assessment index and clinical experience scenarios.
Paired comparison between the clinical experience with non-disabled patients vs. disabled patients
Table 5 compares the means for clinical experience with non-disabled patients (Scenario A) vs. the means for clinical experience with disabled patients (Scenario B). It showed significantly lower means for scenario B individual questions and total score.
Multivariable analysis
A linear regression taking the MSA-PWD scale as dependent variable showed that higher confidence dealing with PWD self-assessment (Beta = 0.705) and Clinical experience with non-disabled patients (Beta = 0.501) were significantly and positively associated with a more positive MSA-PWD (Table 6).
link