Since numerous readers may only peruse the Introduction and Discussion sections, it is critical to organize the report in a way that allows them to quickly locate the areas of interest. The Results section can be seen as the crux of the report, while the Methods section is designed to explain how the results were obtained, and the Discussion provides your interpretation of the results. Some writers prefer to begin with the Results section before moving on to the Methods section. You should follow the order that you feel most comfortable with, but in any case, you will likely need to revisit prior sections to ensure a cohesive report.

Lastly, a crucial general guideline is that shorter is better. If you can accurately convey something in one sentence rather than two, it is preferable. (However, two concise sentences are superior to one long and convoluted sentence.)

The Structure of a Statistical Report

  • Summary
  • Introduction
  • Methods
  • Results
  • Discussion


The Summary section of an analysis report should be self-contained, providing a brief overview of the purpose, methodology, key findings, and interpretation. While it is typically written last, it should use language that is accessible to a general audience. Although it’s preferable to avoid using statistical jargon or concepts in the summary, there may be instances where it is necessary to accurately convey the results. In such cases, it is better to include a slightly technical but accurate summary rather than a simplified one that could be potentially misleading.


In the introductory section, it is important to provide a concise overview of the research question or hypothesis, as well as the context in which the data was gathered and the available data. It should be noted that in this module, the research always begins with a specific problem or question, along with pertinent data.


In the Methods section, it is important to provide a comprehensive description of the model, the procedures used to validate the model, the statistical tests employed, the method used for calculating confidence intervals, and any other relevant techniques used, such as data transformations. The primary objective of this section is to provide enough detail to enable other statisticians to evaluate your approach and replicate your investigation if they have access to the same data. While it is unnecessary to include every failed attempt or dead end encountered during your research (as we all encounter them), if you discovered two equally plausible models that resulted in substantially different outcomes, it would be prudent to include both in this section.


In the Results section, it is essential to include descriptive summaries of the data (such as graphical and numerical summaries), evidence supporting the suitability of your model, and the numerical outcomes of statistical tests or confidence interval calculations. It may be necessary to further round the numerical results when presenting them in this section. However, it is critical to remember that this section, like all others, must be written in prose. Merely presenting a set of graphs and figures is insufficient to convey your findings.


The Discussion section should feature your personal evaluation of the statistical evidence associated with the initial question or hypothesis. Specifically, you should examine any indications of inadequate model fitting, issues with the data (such as outliers), or any other factors that may influence the interpretation of the findings.