The data used came from a survey which took into account 28 different variables, with data being sourced from 2146 students of the DATA1001/1901 cohort (98.9% of cohort).
Specifically, we focused on the following variables: stress level, hours of work per week and the type of student (international or domestic). The variables are classified as quantitative discrete, quantitative continuous and qualitative nominal respectively.
All participants that selected “I DO NOT consent to take part in the study” were removed. Furthermore, data that were determined to be outliers were excluded, making the sample size smaller. Self-reporting bias may have also occurred, as participants could misjudge their stress levels due to perceptional differences. The data also lacks long-term, being cross-sectional it does not account for changes in stress levels or work hours over the semester which could provide more accuracy.
We assumed that all participants understood and attempted all questions truthfully. However, some responses were not reasonable, proven in reporting 999 hours of paid work per week. Therefore we classified work hours exceeding 40 hours per week as outliers. Similarly, regarding stress, we followed the provided scale, treating any response above 10 (worst stress imaginable) as an outlier. We removed these outliers to prevent distortion in our analysis.