Questionnaires and inheritance

The questionnaire has three sections, and they are not interchangeable: each one exists for a different reason.

The three sections

Section 1 — Satisfaction and Section 2 — Climate are the fixed sections. You do not write those questions: they ship with the product, curated, and they are identical for every company measuring with the same theoretical model. That sameness is not convenience — it is the comparability. Two companies on the same model measure exactly the same thing, word for word.

Section 3 is yours. That is where your own questions go, the ones that make sense in your organization and nowhere else. It is the place designed for the particular, precisely so the first two never have to be touched.

That is why Section 3 is editable and the other two are not. If every company could rewrite the system questions, there would be no way to claim two results are comparable, and Section 3 would in effect become the whole instrument.

Inheritance

A questionnaire is assigned to a node, and every descendant inherits it.

That single rule is what makes a two-hundred-node organization manageable: assign it to the root and the whole company has it. If one branch later needs something different —a store with three of its own questions about the night shift— you assign it its own, and that branch stops inheriting from above.

The effective questionnaire

It is the answer to «which questionnaire does this node actually get?»: the one assigned to it if it has one, and otherwise the one from the nearest ancestor that does.

This is the concept that matters when something does not add up. A node rarely has its own questionnaire; what it has is an effective one, which may come from three levels up. If a branch is asking questions you did not expect, the right question is not «which one did I assign?» but «where is it inheriting from?».

Your own questions

In Section 3 you can add questions of six types, and also conditional blocks: a question that, depending on the answer, opens another, and that one another.

Nesting is not capped at one level. A block can go three deep: Are you satisfied? → Why not? → Which part affects you most?

Deleting a question with children takes all of them, at any depth. Before you confirm, the product counts the descendants and tells you — from the list they are invisible, because children display as ordinary questions.

See Question types for the detail on each.

What cannot change, and when

While a study is collecting, its questionnaire is untouchable. Neither the questions nor the assignment of the branch being measured.

This is not administrative rigidity: if the questionnaire changes halfway, the first half of your people answered one thing and the second half another, and both groups get published together as if they were one. The number that comes out means nothing.

For the same reason, if you want to fix a typo in a question, wait until the study closes.

Changing a question can break its series

Editing a question in a way that changes what it asks makes its previous results no longer comparable with the new ones. The question stays alive and the old responses are not lost, but in historical charts the line splits into two segments that do not join.

See The historical series.