Wellbeing questionnaire: the survey, the pitfalls and the blind spot

1. Agreement bias

Scale items with positively phrased statements pull answers upward. "I feel my manager listens" gathers more agreement than reality supports, because agreeing is easier than disagreeing. The effect is well documented and systematic, and it hits management questions hardest.

2. The units are too small

Anonymity requires a threshold. Below five responses a result can in practice point at a person. Many organisations discover late that half their units cannot be reported, and then group them in ways that do not match where decisions are actually made.

3. The average trap

An average of 3.8 can hide a unit where everyone answers 4 and one where half answer 2 and half answer 5. Those are two entirely different problems. Dispersion is often the interesting number, and it is missing from most reports.

4. Vague concepts

"Recognition", "meaning" and "good collaboration" mean different things in a production unit and in a legal department. When a concept is not translated into concrete behaviour, you measure the reading of the word, not the behaviour.

5. The report points at no decision

A report listing ten dimensions with colour codes gives ten possible initiatives and therefore none. Usable reports end in a choice: what do we prioritise, and what do we then deprioritise?

6. The question the survey never asks

None of the six or seven standard dimensions asks: which of our values wins when two of them cannot both be honoured? That question requires a different format — a forced choice in a concrete situation — and that is where a values measurement begins.

See what a forced choice looks like →