[p2p-research] Dunning Kruger Effect (self-assessments of competency)

[Paul D. Fernhout]

Ryan Lanham wrote:
> It would seem this would have a very high import for P2P theory.
> 
> http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect

Of course, there is a complement to this, in that some people who are really 
smart in some area assume everyone around them is just as smart in that 
areas because stuff seems so obvious to them in some problem domain. :-)

For example, to an experienced car mechanic, just listening to a car and 
suggesting fixes is often possible, but to someone else, it may seem like a 
mysterious gift. Genius or intuition often looks like genius or intuition 
only to outsiders, but often represents thousands of hours of experience and 
practice.

Once, almost fifteen years ago when mice were still a fairly new thing for 
many people, we (my wife and I) set up a spreadsheet for a company with some 
important data in it used to configure a larger piece of software. But, we 
had overestimated non-IT people's ability there at the time to be 
comfortable in opening a complex document and not accidentally deleting part 
of it with the mouse without realizing what they were doing. We had thought 
it was a feature that the system was so easily changeable, but, it was 
really a bug in an organizational sense in that context (especially given 
that data had to go through various formal channels to be approved). If the 
spreadsheet had only had textual or numerical data in it, a professional in 
that company might notice stuff like deleting a section, but the spreadsheet 
had code statements as well, so the users had no way of understanding if the 
spreadsheet looked right, another problem. And of course, the non-IT 
professionals did not have version control or difference tools that a 
programmer would have. We made the mistake of assuming that certain non-IT 
professionals had the basic competence of a programmer, even though for us 
at the time, opening a file and being careful in it when mousing around or 
double checking changes was, by then, second nature. So, we failed both in 
underestimating our competency as Information Technology people, and also we 
failed in overestimating our ability to assess the situation and understand 
the limited competency of the otherwise very smart end users in that 
situation (they were also very good with the keyboard, it was just the mouse 
that was the new thing). Ultimately, the end users just asked other 
programmers to maintain that document. :-) So, it might as well have been 
written in code to begin with as part of a larger system. Of course now, 
that most everyone can use a mouse well, that approach might have been a 
little better (although even now I'd hardcode that particular data).

As with your link, people who are inept at a problem domain don't usually 
realize it, as what you suggest, see Kruger and Dunning's "Unskilled and 
Unaware of It: How Difficulties in Recognizing One's Own Incompetence Lead 
to Inflated Self-Assessments"
http://web.archive.org/web/20021112231040/http://www.apa.org/journals/psp/psp7761121.html

But, the paper also suggests competent people tend to underestimate their 
own relative competence, because they typically overestimate the competence 
of their peers, which maybe gives anyone incompetent who reads it another 
reason to inflate their own self assessment. :-)


So, both aspects affect P2P, both overestimating ones competency, and 
underestimating one's competency. And, as with my example above, the two 
issues may be in play with the same people and the same situation all at the 
same time, with someone underestimating one thing, and overestimating another.

--Paul Fernhout
http://www.pdfernhout.net/