Statistical Analysis in a Thesis — A Practical Guide
Statistical analysis in a thesis represents the greatest challenge of the entire writing process for many students. This applies to master's theses as well as bachelor's and engineering theses where the research section is based on quantitative data. Properly conducted statistical analysis not only strengthens the credibility of results but also demonstrates the author's research competence. In this guide, we explain step by step how to approach data analysis in a thesis.
Before you begin analysis, you must plan the research from a statistical perspective. Determine the type of variables (quantitative, qualitative, ordinal), formulate research hypotheses, and select an appropriate sample size. A sample that's too small may fail to detect an existing effect (Type II error), while one that's too large generates unnecessary costs. Statistical power analysis before data collection will help determine the minimum number of observations. It's worth consulting with a statistician or supervisor at this stage to avoid costly methodological errors.
The selection of statistical tests depends on several factors: data type, variable distribution, number of compared groups, and dependencies between observations. For normally distributed data, we use parametric tests — Student's t-test for two groups, ANOVA for multiple groups. When data don't meet normality assumptions, we turn to non-parametric tests: the Mann-Whitney U test, Kruskal-Wallis test, or Wilcoxon test. For qualitative variables, the basic tool is the chi-square test or Fisher's exact test.
Statistical analysis tools are more accessible today than ever before. SPSS remains the most popular program among social science and medical students due to its intuitive graphical interface. R and Python offer greater flexibility and are preferred in exact sciences — R with packages like ggplot2 and dplyr, Python with libraries pandas, scipy, and statsmodels. Statistica and Jamovi are additional alternatives. The choice of tool should be dictated by the specifics of your research and your supervisor's requirements.
Presenting statistical analysis results in a thesis requires following several principles. Always provide the test statistic value, degrees of freedom, p-value, and effect size measure. Tables and charts should be readable, properly labeled, and integrated with the text — don't place a table without referring to it in the chapter content. Describe results objectively in the results section, and leave their interpretation for the discussion. Avoid phrases like "the test showed a tendency" when p > 0.05 — a statistically insignificant result is still a result that needs to be properly interpreted.
The most common errors in statistical analyses of theses include failing to check test assumptions (e.g., distribution normality), applying parametric tests to ordinal data, ignoring multiple comparisons, and confusing correlation with causation. The examination committee and reviewer pay particular attention to these elements. If statistical analysis is not your strong suit, consider consulting with a specialist who can help select appropriate methods and correctly interpret results — it's an investment that pays off in the quality of the entire thesis.