>One one around this is to "Actually have a theory" </TM>
>
>That is, to predict WHY two vars are related. Why should TV and IQ be
>related? Is the relationship causal A->B or B->A?
Theory!!! I once got in trouble in my psychology for arguing that too many studies were fishing expeditions. You go out, trawl the data, find a "significant" difference and publish the results. Theories can just get in the way!
What people don't seem to appreciate is that most constructs like IQ are multifactorial and its almost guaranteed that any two groups are going to differ. Once you find out the difference you can easily create a theory to fit the direction of difference.
People don't even bother to find out if there is a substantive difference. All the fuss about men and women having differing verbal and mathematics ability (based on data from the SAT, and which I believe has now disappeared) was based on a difference of less than 0.5 standard deviations.
>If TV is supposed to lower IQ, you could boot strap off the "music raises
>IQ" data
>(see Rauscher FH, et al. Music and spatial task performance. Nature. 1993
>Oct14;365(6447):611).
>
>However, I have published a pretty nice failure to replicate this study.
>Stough,-Con; Kerkin,-Bridget; Bates,-Tim; Mangan,-Gordon(1994) Music and
>spatial IQ. Personality-and-Individual-Differences.1994 Nov; Vol 17(5):
>695.
>The point is, _theories_ can be tested, relationships cannot.
>
>In psychology this is hardly ever the case. We mostly just have political
>theory wrapped up as science - bull about how people with schizophrenia
>are never violent, IQ isn't heritable, EQ is completely learned, and such
>like.
Couldn't agree more.
>That is why i moved to cognitive science.
Me too. ;^)
ciao, patrick
Patrick Wilken http://www.csse.monash.edu.au/~patrickw/ Editor: PSYCHE: An International Journal of Research on Consciousness Secretary: The Association for the Scientific Study of Consciousness http://psyche.cs.monash.edu.au/ http://www.phil.vt.edu/ASSC/