Likert Scale Calculator
Enter the number of responses at each scale point. Because statisticians genuinely disagree about whether averaging ordinal data is legitimate, this shows both the mean and the median-and-top-box view rather than picking a side.
Free, no signup, and everything is calculated in your browser.
Weighted mean across 100 responses.
Likert data is ordinal, and whether a mean is legitimate is genuinely disputed. Purists report the median, mode, and top-box percentages; practitioners report the mean. Both are shown so you can pick, and so you notice when they disagree.
How Likert results are summarised
Weighted mean = Σ(point × count) ÷ total responses. Median is the middle observation in the cumulative distribution; mode is the most-chosen point.
Top-two-box is the share choosing the two highest points, bottom-two-box the two lowest, and the net is the difference.
The disagreement: Likert points are ordered but not evenly spaced — the distance from "agree" to "strongly agree" isn't necessarily the same as "neutral" to "agree" — so purists report only median, mode and box percentages. Practitioners report the mean anyway because it's easier to track. Watch for cases where the two tell different stories; that usually means a polarised distribution.
Questions people ask
Can I calculate an average of Likert responses?
You can compute one, and most teams do. The caveat is that it assumes the gaps between scale points are equal, which isn't strictly true. If your distribution is polarised — lots of 1s and 5s, few 3s — the mean will hide that and the distribution won't.
Should I use a 5-point or 7-point scale?
5-point is easier to answer and label; 7-point offers more granularity and slightly better statistical properties. Both are defensible. What matters more is labelling every point rather than only the ends.
How do I score a Likert scale?
Assign each point a number (1 to 5 on a five-point agreement scale, with strongly disagree as 1), multiply each point by how many people chose it, add those up, and divide by the total responses. That is the weighted mean. Reverse-code negatively worded statements first — a 5 on "the checkout was confusing" should count as a 1 — or the average of a set of items means nothing.
How do I calculate the mean score for a 4-point Likert scale?
Same method, points 1–4. Because there is no midpoint, the neutral line is 2.5 rather than 3: a mean of 2.7 leans agree, 2.3 leans disagree. Some teams also report it as a weighted mean out of 4 next to the top-two-box percentage, which on a 4-point scale is simply the share who chose 3 or 4.
What is top-two-box?
The percentage of respondents who chose the two most positive points — 4 and 5 on a five-point scale, 6 and 7 on seven. It avoids the equal-spacing assumption behind the mean, is easy to explain to anyone, and is the summary this calculator shows beside the average for that reason.
Further reading
Related tools
Better forms.
Better data.
Build your first form in under 60 seconds. Free forever for personal use, no credit card required.