Joe's Autism Page

those little graphs...

Prof. Simon Baron-Cohen's book, The Essential Difference, contains a half-dozen little graphs. More accurately, perhaps; it contains essentially the same graph, replicated a half-dozen times with different labels on the axes. This is okay – the graph is used to communicate a point to a non-technical audience. However, I believe the graph is used improperly and is not consistent with the author's central argument.

First – some nit picking. The graph is the standard 'normal distribution' (i.e., “bell curve”). The first point to be made is that the empathy quotient (some metric that attempts to describe quantitatively how much empathetic intelligence a person has) is normally distributed. That is:

• most women (or men, but lets examine one gender at a time) have an 'average amount',

• a fewer number have more than average or less than average,

• and a small number have much more or much less than average.

As you move outward from the 'norm' (the center of the distribution) the graph should approach zero asymptotically. None of the author's graphs show this. For some reason he insists on drawing all the graphs with a 'DC offset' – that is, converging on a positive value greater than zero (as shown above). The result seems to say “as you get to the extremes, a substantial residue of people can still be found”.

It would be nice if the author plotted the graphs as shown below, without the 'DC offset'.

End of 'nit'.

Next point... the big one.