Mathemagenic

Via James Robertson - Expertise location without technology by Shawn Callahan. The piece I picked up was on defining expertise:

...expertise is more than simply possessing a skill. Klein describes eight aspects of expertise which I’ve summarised but would recommend you read Klein [Klein, G. 1998. Sources of Power: How People Make Decisions. Cambridge MA: The MIT Press.].

1. Patterns: with experience experts can discern patterns that are invisible to novices. They have a good sense of what’s typical and can therefore detect the extraordinary.
2. Anomalies: experts are surprised when a key event is absent while novices don’t know what is supposed to happen and therefore don’t pick up on the anomaly.
3. The way things work: experts have mental models of how things work—how teams are supposed to work, equipment is supposed to function, power and politics is normally wielded.
4. Opportunities and improvisations: Experts can imagine possibilities that contradict the prevailing viewpoint and data. They can also apply patterns from one context to a new situation creating new approaches and techniques.
5. Past and future: experts can predict what might happen in the future. Just ask a grade 5 teacher about what the kids will be like at the beginning and the end of the year.
6. Fine discriminations: experts can see differences which remain invisible to novices. Just think of expert wine tasters.
7. Self aware: experts are aware of their own thought processes.
8. Decision makers: experts can make decisions under time pressure.

Which in a funny way connects to my thinking of researcher's role in research - for example, differences that would emerge if a particular dataset is analysed by novice vs. expert.

And it comes back to my long-time burning question - what is methodologically sound way for recognising patterns, anomalies, opportunities, fine discriminations in an expert way?

If expertise is difficult to articulate, how would you specify (for example) explicit coding criteria to pinpoint patterns? How far the need to make things explicit, to categorise beforehand would ruin the richness of what could be found? How far the decisions on what are the patterns could be logically explained? How easily the process itself could be articulated for an examination by others?

How the world full of complexity and emergent things could be simplified to a clean-and-clear logic of a methodologically sound process?

Thinking of Making a Mess with Method by John Law and wondering why the hell I can't do something easy - focusing on content instead of methodology... I guess I'm still in search of that particular messy method that fits the way I deal with the world and of a scientific environment where I don't have to defend it...

Something to read for those seriousely looking at autoethnography for their research - Representation, Legitimation, and Autoethnography: An Autoethnographic Writing Story by Nicholas L. Holt

Abstract: The purpose of this article is to critique representation and legitimation as they relate to the peer review process for an autoethnographic manuscript. Using a conversation derived from seven reviewers’ comments pertaining to one autoethnographic manuscript, issues relating to (a) the use of verification strategies in autoethnographic studies; and, (b) the use of self as the only data source are discussed. As such, this paper can be considered as an autoethnographic writing story. The problematic nature of autoethnography, which is located at the boundaries of scientific research, is examined by linking the author’s experiences of the review process with dominant research perspectives. Suggestions for investigators wishing to produce autoethnographic accounts are outlined along with a call for the development of appropriate evaluative criteria for such work.

Make sure you check references as well.