----- Original Message ----- From: "Mike Palij" <[email protected]>
To: "Teaching in the Psychological Sciences (TIPS)"
<[email protected]>
Cc: "Mike Palij" <[email protected]>
Sent: Monday, October 05, 2009 6:13 PM
Subject: re: [tips] Anxiety article in NYTimes Magazine


On Mon, 05 Oct 2009 14:06:39 -0700, Beth Benoit wrote:
Sunday's NYTimes did a nice job explaining Jerome Kagan's
research on temperament.  Most of the research articles we read
don't give little tidbits of interesting information, and I find these
are helpful to spiff up my lectures in class.
http://www.nytimes.com/2009/10/04/magazine/04anxiety-t.html

I admit that it is a good read but one has to be cautious about
drawing easy conclusions from the decades of research Kagan
and colleagues have performed.  The most obvious problem is
that of sampling bias.  As the article mentions, most if not all
of the subjects were white, middle-class, and born healthy.
Not mentioned is "what kind of parent allows their 4 month old
participate in a research study?"  Surely such parents would be
different from parents who would refuse to have their children
participate (much like volunteer subjects differ from non-volunteers).
Self-selection for follow-up participation further biases the
sample.

In the end, to whom do these results generalize to?  Do they
tell us anything about the temperament of non-white,
non-middle-class, and infants born with health difficulties?

Perhaps Kagan's research has far less generalizability then he
and his colleagues are will to admit, far less than that implied
in the article.

-Mike Palij
New York University
[email protected]

Mike: I do not not see sampling bias as an issue here. All what the authoers
have to say that these studies invovled groups with certain characteristics
and should be cautious in generalizing to other groups.Researchers should be
afraid to atate "we do not know since we did not test for it." There is
nothing wrong in doing a study with all whites with a certain income
level.What we do not want to do is to generalize to other groups when we
have not ascertained the same research on other groups. I guess that is
where designs should always test for interactions.

I personally do not care for research on temperament.I have had nurses tell
me that they observed higher activity level for black babies than white
babies.How they measure that I do not know.Besides temperament is a
descriptive category.To me real science has to do with explanations.

Michael Sylvester,PhD
Daytona Beach,Florida


---
To make changes to your subscription contact:

Bill Southerly ([email protected])

Reply via email to