What it does
Likert questions (“Strongly disagree” to “Strongly agree”) are the backbone of staff surveys, course evaluations and social research. The clearest way to show them is a diverging stacked bar chart: each statement is one bar, disagreeing answers stretch to the left of a centre line, agreeing answers to the right, and the neutral answer sits across the line. Statements people agree with stand out on the right at a glance.
This tool makes that chart from your data, in your browser:
- from raw answers (one row per person, one column per statement), as codes 1 to k or as the answer labels, or from a table of counts per answer;
- with 4, 5, 6 or 7 answer options and your own labels;
- sorted by agreement, by disagreement, by mean, or kept in your order;
- with a summary table: number of answers, % disagree, % neutral, % agree, median and mean (SD) for each statement.
Download the chart as PNG or SVG (vector, for reports and slides), the summary as CSV, or everything as an image, PDF or Word report.
It works well with the reliability and agreement, the publication plot maker and the sample size and power calculator.
How to use it
- Paste your answers or open a CSV or Excel file. Each column header is the statement text. Columns such as an ID are left unticked; tick or untick the statements you want.
- If you only have totals, choose Counts per answer and paste one row per statement with its counts, most negative first.
- Choose the number of answer options and edit the labels if yours differ (separate them with commas). Answers given as labels are matched in any letter case.
- Pick the order and, if you like, a chart title. Blank answers are skipped.
Worked example (made-up data)
The sample is a made-up staff survey: 150 people rated 8 statements from 1 (strongly disagree) to 5 (strongly agree). Sorted by agreement, “I feel part of a team” comes first with 72.0 % agree, 7.3 % disagree and a median of 4. At the bottom are “Pay is fair for my role” (22.3 % agree, 58.8 % disagree, median 2, from 148 answers because two were blank) and “Meetings are a good use of time” (20.0 % agree, median 3). “I have the tools I need” splits opinion: 50.3 % agree but 34.9 % disagree.
How it was checked
The counts, percentages, % agree and % disagree, medians (numpy’s median, which averages the two middle answers when the number of answers is even), means and standard deviations (pandas, n − 1) agree with Python for every statement in the example. The chart layout is checked so that the bars of each statement touch, add up to 100 % and put the neutral answer exactly across the centre line.
Tips
- Report the percentages and the median. Likert answers are ordered categories; the mean treats the gaps between them as equal, which they may not be.
- Keep the same scale for every statement. Reverse-worded statements should be recoded first so that “agree” always means the favourable answer.
- Look at the neutral share. A statement with few disagreements but a large neutral middle is not the same as one people actively endorse.
Frequently asked questions
Can I use a scale without a middle option? Yes: choose 4 or 6 answer options and the bars split at the centre line between the last disagreeing and the first agreeing answer.
Is my survey data uploaded? No. Everything runs in your browser and the page keeps working offline.
Where the numbers come from
- Method: Likert (1932) rating scales; diverging stacked bar charts as described by Robbins and Heiberger (2011). Example data made up.
Standard scientific values, such as constants, atomic weights and melting points, are facts found in every textbook; we name where ours were checked. If you are an author or publisher and want a value, a credit or a table changed or removed, tell us and we will act quickly.
Last updated 2026-11-09.