IRBs should focus on assessing the potential risk for harm to
participants and should not address psychometric issues of the study. I
believe this for a few reasons.

First, an IRB cannot make an educated decision on psychometric issues if
they do not have expertise in that particular content area. I may decide
to use a measure that doesn't yield very reliable scores, but may be the
best measure available. Plus, as the primary investigator for a research
study it is my responsibilty to choose the measures. If I choose
measures that are poorly constructed or do not produce reliable/valid
scores, then the editors reviewing my paper for publication will
(hopefully) catch it.

Second, if the IRB is going to evaluate score reliability, then at what
cut-off point are they going to decide that an instrument poses a risk?
Are they going to go by the .70 criteria? Higher? Lower? This is a
slipperly slope that is best avoided.

Third, technically reliability is a property of scores and is not a
property of tests themselves. When tests are developed, they do not have
a reliability coefficient stamped upon them by the almighty publisher.
Researchers should ALWAYS calculate score reliability and validty with
their current samples and not rely on previous estimates from other
samples. In fact, many journal editors are now requiring researchers to
do this prior to submitting articles for publication.  

The argument that participants need to be protected from the potential
risk of wasting their time completing surveys that do not provide
reliable scores is a weak argument. Maybe we should not let students
complete paper surveys because they will run the risk of getting paper
cuts. But we couldn't have computer surveys because participants may be
at risk of developing carpal tunnel syndrome. Maybe we should require
researchers to write their surveys backwards so left-handed participants
won't run the risk of smearing their answers and getting ink on the
hand. Okay, so I know I'm being ridiculous here (it's the day before
grades are due!). This whole situation has too much micro-management and
I wonder what kind of political factors are playing into their decision.

Rod

______________________________________________
Roderick D. Hetzel, Ph.D.
Department of Psychology
LeTourneau University
Post Office Box 7001
2100 South Mobberly Avenue
Longview, Texas  75607-7001
 
Office:   Education Center 218
Phone:    903-233-3893
Fax:      903-233-3851
Email:    [EMAIL PROTECTED]


---
You are currently subscribed to tips as: [EMAIL PROTECTED]
To unsubscribe send a blank email to [EMAIL PROTECTED]

Reply via email to