Data Analysis and Computational Applications
Core concepts
Candidates should understand the data-analysis pipeline: collection, cleaning, transformation, analysis, visualisation, and interpretation. Know data types (categorical, ordinal, discrete/continuous numeric), sources of bias (sampling, response, confirmation), and the difference between correlation and causation. Be able to read tables and charts, compute simple statistics (mean, median, mode, range), and choose appropriate visualisations: bar charts for categories, histograms for distributions, scatter plots for relationships, line graphs for trends over time. Understand basic database queries (SELECT, WHERE, ORDER BY), spreadsheets (formulas, absolute vs relative references), and how computational models (simulation, machine learning) are applied in science, medicine, and business.
Traps
- Mean is skewed by outliers; median is more robust-questions test which to report.
- A histogram shows frequency distribution; a bar chart compares categories-swapping them is a common distractor.
- Correlation does not imply causation; a lurking variable may explain both.
- Relative cell references shift when copied; absolute ($A$1) do not.
- A biased sample cannot be fixed by increasing sample size.
Worked example
Scores: 4, 6, 6, 7, 42. Mean = 65/5 = 13; median = 6; mode = 6. The mean is distorted by the outlier 42, so the median better represents typical performance. Rule of thumb: if the distribution is skewed or contains outliers, report the median; if it is symmetric, the mean is informative.