Yes, I goofed. 'Confounds' are generally used in the sense you describe
below. The problem I described has mor to do with external validity.
Jim Clark wrote:
>
> Hi
>
> On Thu, 1 Apr 1999, John W. Kulig wrote:
> > Spotting confounds is best done after you know what the researchers
> > concluded. _If_ the researcher were to claim that men, in general, are more
> > helpful, we'd have a task confound because the study used only one task -
> > helping with keys. Ideally, you'd want to utilize a variety of tasks in
> > different settings, and, you might see a task-by-gender interaction. On the
> > other hand, if the researcher merely claimed that men are more helpful in this
> > situation, tbe criticism disappears.
>
> I would not call this a confound. Confounds generally concern
> internal validity; that is, some factor that is correlated with
> the factor under study (in this case gender). Task cannot be
> correlated with gender because it has no variation. The problem
> of task appears to me to be associated with external validity;
> that is, are the results generalizable to other tasks/situations?
> As John notes, we could be more confident of this if multiple
> situations had been used.
--
* John W. Kulig, Department of Psychology ************************
* Plymouth State College Plymouth NH 03264 *
* [EMAIL PROTECTED] http://oz.plymouth.edu/~kulig *
*******************************************************************
* "Eat bread and salt and speak the truth" Russian proverb *
*******************************************************************
