Accountability and disclaimer

The basis

Where the method comes from

Group Concept Mapping was described in the 1980s by William Trochim and later developed further with Mary Kane. The method is published and free to use. ConceptAtlas is an independent implementation of it and uses no code from other providers.

From sortings to a map

Each participant places statements into piles. How often two statements end up in the same pile determines how closely they sit on the map. That computation is called multidimensional scaling.

The themes then emerge through clustering. The names are given by the group, not by the algorithm.

How reliable is that map

Stress indicates how faithfully the flat map represents the original sortings. For this type of map a value between 15 and 35 is customary.

Split-half reliability splits participants randomly into two halves and compares both maps. Above 0.80 counts as sound. That tells you whether the outcome depends on who happened to take part.

What bridging adds

Bridging shows whether participants agree on where a statement belongs. A low value means a sharply delineated theme; a high value means people think differently about it.

That is a finding in its own right: where there is no shared understanding, there is little to decide yet.

What the two scales measure

Importance and feasibility are not simply two opinions. They map onto the behavioural theory of Fishbein and Ajzen: importance corresponds to the evaluation of the outcome, feasibility to what Ajzen calls perceived behavioural control.

That is more than a footnote. Perceived control is the stronger predictor of whether people actually act. If a theme scores high on importance and low on feasibility, stressing its importance will not help; increasing control will.

What the simulation does

The modelling is based on fuzzy cognitive mapping. The group itself judges how strongly themes influence one another; those judgements become the weights in the model.

The result is an exploration of direction and dynamics. It is explicitly not a prediction, because the weights are opinions rather than measured effects.

Disclaimer

What ConceptAtlas does and does not claim

A shared map is a powerful instrument, but it remains an instrument. These are the limits, stated explicitly.

Scores are judgements, not measurements. Impact and feasibility say what participants expect and consider important. They do not say what demonstrably works.
The simulation predicts nothing. It explores how a group model behaves when you push on a theme. Outcomes are not a forecast of reality and should not be presented as one.
The map does not replace the conversation. The analysis provides input for a working session. Interpretation and decisions remain with the group.
The AI proposes, it does not pronounce. The assistant suggests cluster names and phrasings based on anonymous content. A human always decides; nothing is adopted automatically.
An outcome holds for this group, at this moment. Results do not automatically generalise to other organisations or to later points in time.
The sample report contains simulated data. It shows the shape of the deliverable, not the outcome of a study that was carried out.
Not advice that replaces professional judgement. We provide an evidence base for decision-making. Responsibility for the decision rests with the client.

Sources

Where you can look it up

The method and the benchmarks used are published. Below is which source was used for what, so you can trace every claim.

Used for	Source
The method itself	Kane, M., en Trochim, W. (2007). Concept Mapping for Planning and Evaluation. Sage.
Original description	Trochim, W. (1989). An introduction to concept mapping for planning and evaluation. Evaluation and Program Planning.
Benchmarks for stress and reliability	Rosas, S. R., en Kane, M. (2012). Quality and rigor of the concept mapping methodology: a pooled study analysis. Evaluation and Program Planning.
Meaning of the rating scales	Fishbein, M., en Ajzen, I. (1975). Belief, Attitude, Intention and Behavior. Addison-Wesley. · Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes.
The simulation	Kosko, B. (1986). Fuzzy cognitive maps. International Journal of Man-Machine Studies.
From map to design	McKenney, S., en Reeves, T. C. (2018). Conducting Educational Design Research. Routledge. · Sandoval, W. (2014). Conjecture mapping. Journal of the Learning Sciences. · Van den Akker, J. (1999). Principles and methods of development research.

The exact algorithms, the choices made and the licence of the software used are described openly on the method page. If you work as a researcher, see also from cluster map to design.

What this rests on, and what it is not