Guide
Five ways to report Likert course evaluations without the mean
Better summaries for ordinal rating scales
Five practical alternatives to reporting a mean score for Likert-style course evaluation items, which are ordinal rather than interval data. Intended for institutional researchers, quality teams and committees that compare courses or teachers on rating scales. Koji supports this by pairing structured questions with conversational interviews and automatic thematic analysis, so decisions do not rest on one average.
What's in this guide
- Report the full distribution: Show how many students chose each option. Two courses can share an average and tell opposite stories.
- Use the median and mode: They suit ordinal data and are honest about small differences.
- Use rank-based tests: Compare groups with methods built for ordered categories, not t-tests on codes.
- Stop ranking to the decimal: If two responses could flip the order, it is noise with a number attached.
- Give weight to written feedback: It explains why a course worked or did not, which a score cannot.
Use this image
Free to use in your docs, decks and posts under CC BY 4.0, as long as you credit Koji with a link to this page. To embed it, paste this into your page:
html
<a href="https://www.koji.so/visuals/reporting-likert-course-evaluation-without-the-mean"><img src="https://www.koji.so/visuals/reporting-likert-course-evaluation-without-the-mean/reporting-likert-course-evaluation-without-the-mean.png" alt="List of five ways to report Likert course evaluation data without averaging, from showing distributions and medians to giving comments more weight." width="1200" height="790" /></a>
<p>Source: <a href="https://www.koji.so/visuals/reporting-likert-course-evaluation-without-the-mean">Five ways to report Likert course evaluations without the mean</a>, Koji</p>