Rigorous idiography

Page created 4.i.19. All content is made available under a Creative Commons License. Please feel free to reuse anything here but respect the licence, i.e. give attribution back to here. 

“Rigorous idiography” is a term I’ve been using for probably a decade or more now.  My paper with John Hughes and Julia Houston (Evans, Hughes, & Houston (2002) Significance testing the validity of ideographic methods: a little derangement goes a long way. British  Journal of  Mathematical and Statistical Psychology, 55(2), 385–390. https://doi.org/10.1348/000711002760554525) was my first in this area and I don’t think I’d hit on the generic term back then.  The idea is that there are forms of data that are purely idiographic: a picture, description or summary that is unique to, and of, an individual, but that this absence of any simple dimension on which the data can be ranked or rated against that of any other individual doesn’t and mustn’t preclude us doing things rigorously with such data where we can.  

One now completely accepted example, though not quite what I’m talking about, is analysis of fingerprint data: the patterns are essentially unique to each of us and we have good descriptive terms to classify them, but the differences between them don’t reduce to a few dimensions on which they can be compared. Despite this the rigorous ways to describe them allows them to be used to identify individuals.  A more recent and similar example is the genome: we each have a unique one and there is no single dimension that compares any one with any other, nor even a small set of such dimensions.  However, we can describe them completely rigorously and identify differences between people on them.

I have no wish to see qualitative data reduced by quantification where it’s not appropriate or helpful and that would probably apply to most therapy narrative data. However, there are data that are purely idiographic but which can still be susceptible to rigorous mathematical analyses that help strengthen the argument that the data contains important and meaningful information. This is the domain I am calling “rigorous idiography”. 

To some extent the mathematical analysis of repertory grid data is one such method and it happened that we came up with the method of derangements out of some repertory grid data but it’s applicable to any descriptive data about a set of individuals. That a good set of fingerprints from a set of people, or DNA sequencing from them, will distinguish between those people is one, rather extreme, example of the method of derangements.  I’m keen to see it used for a wide variety of qualitative descriptors as I think it will help dent the widespread quantitative researchers’ denigration of such data.

There are other good tools like this, the method of mismatched cases is one that has had a small amount of use in our area.  I’m clear this is an area that could be developed with huge impact.