It is imaginary data - I'm just looking for the appropriate shape. I'm using the following for my bell curve:

http://classes.kumc.edu/sah/resource...bell_curve.gif

(Found off the net) and covering up the arrows and 'more than usual' by putting yellowish coloured rectangles over them.

Now what I was looking for was a bell curve with chunks cut out of the left. If you open up the attachment that is with my very first post to the thread hopefully that would make it clearer...

Ideally... I'd have a bell curve... then a bell curve with chunks out of it... In the same format with the same colour etc etc etc. Preferably black background with white lines... But I'm crap crap crappy with these things.

(The idea is basically... If we imagine (this is probably false) that people have varying abilities to focus their attention and that the variance in peoples ability falls into a normal distribution... Then we want to know 'where do we draw the line on how severe you need to be in order to qualify as having ADHD? If we draw the line at 2 standard deviations below the mean then we fix prevalence at 2%. If we draw the line at 1 standard deviation below the mean then we fix prevalence at 16%. But how do we decide where to draw the line?

If we (rather simplisically) Imagine that the amount of dopamine present in the synaptic cleft of different individuals similarly has a normal distribution... And we grant the (oversimplistic) assumption that 'too little dopamine causes ADHD' then the problem recurrs: How little dopamine qualifies someone as having ADHD? If we draw the line at 2 standard deviations below the mean then we fix prevalence at 2%. If we draw the line at 1 standard deviations below the mean then we fix prevalence at 16%. How do we decide?

Then the idea is that it might be that there isn't a normal distribution (of either ability to focus attention, or of amount of dopamine) after all. It might be that there are 'plateaus' where a certain number of people are clustered at a certain ability (or disability) point. Does that give us reason to believe that they have a dysfunction? Not sure... But thats what I want that for at any rate. I've heard... Though maybe this is controversial that we find those plateaus in IQ as well and that is why we think that certain people are disordered compared with just being at the low end of the normal range. But I've googled 'IQ graph' and stuff like that and I'm not finding a normal distribution curve with plateaus / chunks cut out of it.

Of course we can run through the whole thing again with serotonin / depression and dopamine / psychosis as well.

And since the medications affect ALL people similarly (who doesn't find their cognitive performance improves on low levels of amphetamine???) the effectiveness of medication doesn't help us settle this issue either.

The issue is basically... What is it about a person that determines that they have a disorder compared with being within normal range? I'm trying to say... That even when the neuroscience comes in we still won't know the answer to that so we better start looking at our values...

And... If too many people take 'performance enhancing' medication then it will be a fitness trap (need to take it to keep up and the cost of defecting will be to high.) Only giving meds to people who are 'disordered' stops it being a fitness trap... But where do we draw the line on who is disordered or not? Is it fairly arbitrary or is there a more systematic distinction than that?

So.. The actual data doesn't matter much. And I'm hoping that whether or not attention (or depression or psychosis) falls on a bell curve or not... Whether serotonin or dopamine or whatever falls on a bell curve or not... Whether serotonin or dopamine or whatever actually causes the disorder or not (I know that is far too simple) that it will be enough to illustrate teh basic conceptual point?? - that neuroscience won't tell us whether someone is disordered or not.)