Why do the percentages not add up to 100% in the distribution of:
In the distribution of closed questions
You are analyzing the distribution of your closed-ended questions and you can't explain a 1% anomaly? If you group all categories of respondents, you can get 101% or 99% of respondents instead of 100%.
This slight discrepancy can be explained simply by the use of whole numbers. To make it easier to read and use the results, all percentages are rounded to the nearest whole number.
For example, in the breakdown below, when all categories are added together, the result is 101% of respondents.
Catégorie de répondants | Actual percentages | Rounded results |
Positive | 50,61 % | 51 % |
Neutral | 42,2312 % | 42 % |
Negative | 4,51283 % | 5 % |
Strongly negative | 2,64597 % | 3 % |
Why this choice?
Rounding avoids displaying percentages with several decimal places, such as "33.56789%" which in the interface appears as "34%". This way you can see the main trends at a glance.
Don't worry: with a maximum of 1% accuracy, all Supermood results are extremely reliable, these variations are sometimes disturbing... but negligible!
In the distribution of multiple-choice questions
If employees can check only one answer: the explanation is the same as above, it's a question of rounding!
On the other hand, if employees are allowed to check more than one answer for the same question, the total number of answers very quickly exceeds the number of respondents. When you add up the percentages, you end up with a total that is well over 100% - and that's normal. This format mainly allows you to analyze trends.
If you prefer results that allow you to categorize populations, you can turn to single-choice MCQs.
→ For more details: Composing a Multiple Choice Question (MCQ)
For any other questions, drop us a line at [email protected] - we'd love to hear from you. 💁♀️