Similarity can be misleading

In a recent series of posts ([3], [4], [5]) on similarity in cognitive maps and cause maps I outlined the importance of similarity. I wrote that similarity is one way to find undiscovered links. These new links, in turn, would show where participants (of the Causal Cognitive Mapping exercise) have common grounds. That is important, do not get me wrong. But an overemphasis may be misleading.

Misleading

First of all, the kind of similarity I was talking about was pairwise similarity. That means I was only interested in comparing two statements at a time and judging their similarity. However, there can as well be similarity between one statement and a group (or set) of statements and between groups of statements ([6], p12).


There is also a very practical philosophical issue pertaining to the symmetry of relationships ([7], p137). The measures I have mentioned treat relations in that A->B = B->A. This is true only if A and B can be merged. Otherwise there is a direction of causation the mapper has to assign.


Stressing the importance of similarity can be misleading too. What about dissimilarity? Why do parts not match? Could there be a statement that can connect dissimilar parts? Why not try to make a similar set of statements dissimilar by removing statements and thereby investigate how resilient the set is to "dissimilarity shocks".


Another misleading implication can be if similarity is considered solely as something good. Of course, "[i]f maps are the same across several persons then it is possible to conclude that each person construes the world in the same way as another" ([1], p314). So "Yay, no conflict"? Well, similarity in cause maps may indicate groupthink ([2], p361). That is the unhealthy suppression of conflict and everything but good. [8]


Finally, cause maps are often dynamic which suggests that an investigation into how similarity came about is equally important (p361-2). An apparent similar pair of statements may very well turn out to be emergent from different, even conflicting, explanations (p362); particularly important for groupthink issues.


(C) CC BY (https://creativecommons.org/licenses/by/4.0/), Jo. Richter, http://causal-cognitive-mapping.blogspot.com/2017/02/similarity-can-be-misleading.html


[1] Ackermann, F. and Eden, C. (1994), "Issues in computer and non-computer supported GDSSs", Decision Support Systems, No. 12, pp. 381–390.
[2] Ackermann, F. and Eden, C. (2005), "Using Causal Mapping with Group Support Systems to Elicit an Understanding of Failure in Complex Projects: Some Implications for Organizational Research", Group Decision & Negotiation, Vol. 14 No. 5, pp. 355–376.
[3] http://causal-cognitive-mapping.blogspot.de/2017/01/surprising-finding-similarity-in.html
[4] http://causal-cognitive-mapping.blogspot.de/2017/01/the-tragedy-of-similarity-in-cognitive.html
[5] http://causal-cognitive-mapping.blogspot.de/2017/01/dealing-with-uncertainty-of-similarity.html
[6] Everitt, B. (1980), "Cluster Analysis", 2nd Ed, Heinemann Educational Books Ltd., London.
[7] Uprichard, E. (2009) "7. Introducing Cluster Analysis: What Can It Teach Us about the Case?" in Byrne, D. and Ragin, C.C. (eds), The SAGE Handbook of Case-Based Methods, SAGE Publications Ltd, London.
[8] Just in case this confused you. Yes, conflict is necessary. It is a bit like your body that needs to get ill sometimes in order to train the immune system - occasionally even deliberately in case of vaccination. In a way you can say that the absence of conflict furthers entropy.