Objectives: Medical students require a solid understanding of statistics in research papers. However, literature lacks sufficient studies on their proficiency, especially on Jordanian aspect. Methods: We performed a cross-sectional study on 361 medical students to evaluate their attitudes toward statistics, their knowledge in research methodology, and their ability to correctly interpret commonly used statistics in medical research. Statistical attitudes were assessed using three scales (value, difficulty, and interest) from the SATS-36 scale. Additionally, a novel author's-validated scale was created to measure students' research methodology knowledge. The scale validity was assessed using Confirmatory Factor Analysis (CFA). Multiple linear regression analysis was used to define predictors of statistical inference knowledge among medical students. Results: Of participants, 52.4% were females and 68.2% were in the clinical stage. Research methodology knowledge scale showed adequate model fit indices (CFI, TLI, NNFI, and GFI all = 0.98, with RMSEA value of 0.06 and high reliability (Cronbach’s alpha = 0.89)). Positive attitudes towards the value of statistics (Coefficient = 0.62; 95% Cl: 0.17-1.06; P = 0.007) significantly predicted better statistical inference knowledge. In addition, lower perceived difficulty was the strongest predictor of knowledge of statistics (coefficient = 1.32; 95% Cl: 0.62-2.03; P < 0.001). Significant correlations were also found with higher academic performance and years of study (P < 0.05). Additionally, the cross-sectional design and simple random sampling were the most familiar study designs and sampling methods, respectively. Notably, 49% of students reported no prior knowledge of any statistical test, while the chi-square test was the most known one (36.3%). The predominant method for learning research methodology was through online videos and websites (76.6%), and SPSS was the most known statistical software (58.7%). Conclusion: Our findings may improve medical students' statistical inference skills in medical research. Furthermore, our assessment scale for research methodology knowledge will assist researchers evaluating such knowledge in various contexts and other populations of interest.
Ashour et al. (Thu,) studied this question.