Making differences between groups visible

2026-07-15

In almost every organisation there are people with different positions at the table. Practitioners, managers, support staff, sometimes clients or students. They discuss the same subject and assume they are talking about the same thing.

Usually that is only half true.

One map, several sets of scores

A useful property of concept mapping: the map itself is made by the whole group together, but the ratings can be calculated separately per subgroup. That way you keep one shared structure and still see where the groups diverge.

This is deliberate. If you made a separate map per group, you could no longer compare them; you would be comparing apples with a different arrangement of pears.

What you typically see

There is one pattern worth watching for: a difference in feasibility, not in importance. It may be that practitioners and administrators consider the same things important and differ only in judging whether it will work. Whether that holds has to emerge per project; comparing roles is, however, a standard move in the literature.

That difference is diagnostic. If the shop floor considers a theme far less feasible than the leadership does, it is rarely about unwillingness. Usually it is about preconditions that are invisible from above: time, timetables, authority, equipment that does not work.

Where importance is equal and feasibility diverges, you have a preconditions problem, not a motivation problem.

How to present it

The usual display is a graph with the theme scores of group A against those of group B. Themes on the diagonal are rated equally; themes far off it are what you talk about.

A correlation figure belongs with it, indicating how strongly the two groups align in general. A high correlation with two outliers is a very different story from a low correlation across the board, and that distinction is lost if you only look at the picture.

Two pitfalls

Subgroups that are too small. Below eight participants per group such a comparison becomes unreliable; one deviating participant then skews the picture.

Traceability. With small groups a breakdown can make individuals identifiable, certainly if only one administrator took part. Then you should not make that breakdown, however interesting it is.

How to set up such a comparison is shown under The process.

← All articles