Hi On 6 Dec 2001, David Heiser wrote: > Most of the focus is on structural equation modeling (SEM). For > statisticians, a quick referral to Jim Steiger's article "Driving Fast in > Reverse" in JASA March 2001, p331-p338 (if you have it around) is a quick > discourse on SEM and the inherent problems of figuring out what is going on > from a model (I can send a copy via e-mail attachment if anyone asks)..
Perhaps it is the empiricist in me (or the lack of mathematical sophistication), but I would rather trust to more data to help track down causal hypotheses in non-experimental data. As a prototype, I like Anne Streisguth's continuous efforts to show that no matter what additional variables are measured and included in a multiple regression (20 or more in some older studies I have seen, so perhaps even more now), there continues to be an effect on childhood intelligence (or some other outcome variable) of maternal alcohol consumption that is independent of all the other predictors. The best way to bolster claims about causal relationships is to eliminate through empirical study competing hypotheses. This can most quickly and easily be done when the topic lends itself to experimental manipulation (e.g., effect of maternal alcohol consumption in randomized animal studies), but it can still be achieved (or approximated) non-experimentally, albeit in a more laborious and gradual manner. Is it not the case that many of science's greatest models (e.g., evolution, continental drift) were arrived at by explaining non-experimental data that were problematic for other models? Best wishes Jim ============================================================================ James M. Clark (204) 786-9757 Department of Psychology (204) 774-4134 Fax University of Winnipeg 4L05D Winnipeg, Manitoba R3B 2E9 [EMAIL PROTECTED] CANADA http://www.uwinnipeg.ca/~clark ============================================================================ ================================================================= Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =================================================================
