On Thu, September 30, 2010 11:31 am, Kyle Gorman wrote:
> I agree with Josef. To borrow a tortuous analogy from Baayen, I'd rather
ride in a car driven by a rube and built by an engineer than a car
driven
> and built by a hobbyist.

Maybe, but having led campaigns for safer streets, let me say that the
analogy is a good one, but it supports the opposite conclusion of the one
you reached.  I'd rather be crossing the street in front of a car built
and driven by a hobbyist, because it's obvious to everyone (including the
hobbyist, we hope) that extra care needs to be taken.

I lost my high school history teacher to a rube driving a car built by
engineers.  The engineers gave the rube a false sense of confidence that
allowed him to get behind the wheel after a night of drinking.  Years
later, I saw a man die in front of my eyes because someone operated his
high-tech device while feeling sick.

So it is with statistical tests.  SPSS and SAS look so smooth, researchers
think that all they have to do is plug their numbers in and out comes the
answer they want.  And so they do, regardless of whether the relevant
assumptions apply.

It's true that some experiments require complex models, but those should
at least be reviewed by people who understand those models.  Slick
packages can give a misleading impression.

For example, as I wrote in the paper below, Biber's 1988 book is a factor
analysis based on the assumption of independent variables.  The assumption
is not valid in this case, so the results are problematic.  It's a complex
model that was applied without proper care, with misleading results.

http://www.grieve-smith.com/Academic/AAACL-grvsmth.060225.pdf

--
			-Angus B. Grieve-Smith
			[log in to unmask]

########################################################################

The Variationist List - discussion of everything related to variationist sociolinguistics.

To send messages to the VAR-L list (subscribers only), write to:
[log in to unmask]

To unsubscribe from the VAR-L list, click the following link:
http://jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=VAR-L&A=1