I understand that Likert scales are often used in survey and marketing research
that is purely correlational in nature. However, isn't there a more general
methodological issue here as well? If one uses a truly experimental design,
people's interpretation of the labels would be obviously be randomly distributed
among groups, and the effects would thus wash out. But, as you all know, in
performing studies like those involving gender differences, we don't have that
luxury. So, wouldn't we have a similar problem in making assertions about gender
differences, regardless of whether we used a Likert-type scale, a multiple-choice
survey, an open-ended survey, etc? After all, there may very well be differences
between genders in how they interpret the question itself, as well as the labels
on the Likert scale. As such, while I understand the logic of not making
between-group comparisons with Likert scales, I'm not sure that the same logic
cannot be applied to the use of other survey methods. Any thoughts on this would
be appreciated.
Take Care!
"Kenneth M. Steele" wrote:
> On Tue, 24 Oct 2000 08:37:10 -0400 "John W. Kulig"
> <[EMAIL PROTECTED]> wrote:
>
> > Supposedly, it is tricky (or, downright wrong) to use these scales for
> > across group comparisons, they are only for within-group comparisons. The
> > reason is that different groups' use of the adjectives may not be
> > equivalent. If we compare "tasters" versus "nontasters" for instance, the
> > the anchor "extremely bitter" for the former may correspond to a very
> > intense stimulus - subjectively as intense as a very loud noises. For the
> > latter, an "extremely bitter" taste may be at the top of gustatory
> > experiences - but relatively mild relative to very loud noises. That is, the
> > sensory world of tasters vs. nontasters guarantees these adjectives will be
> > interpreted differently. The same logic would apply to other between-group
> > comparisons.
>
> I agree, but the major problem here is caused by comparing
> across nonequivalent groups.
>
> I see this problem ignored most often when people are talking
> about "gender differences." Consider the following
> stem-and-leaf plots of answers to the question---
>
> Is "Doom" a gory computer game?
>
> 1 2 3 4 5 6 7
> Not Very Medium Very
>
> Men Women
>
> 7 7 +++
> 6 +++ 6 +++++++
> 5 +++++++ 5 ++++++++++
> 4 ++++++++++ 4 ++++++
> 3 ++++++ 3 ++
> 2 ++ 2 +
> 1 + 1
>
> Notice that the distribution of Men's answers are shifted
> towards "Not" by 1 item but are otherwise identical to the
> Women's answers.
>
> It doesn't matter whether you treat the data as interval or
> ordinal (the usual argument involving Likert-type scales). The
> summary statistic would indicate that men rate the game as less
> gory.
>
> As John indicated, the problem with this conclusion is that we
> don't know whether males and females use the scales in the same
> manner. The difference may only mean that men and women have a
> systematically different definition of the words "not very."
>
> Ken
>
> ----------------------
> Kenneth M. Steele [EMAIL PROTECTED]
> Dept. of Psychology
> Appalachian State University
> Boone, NC 28608
> USA
--
****************************************************
Steve Vanden Avond, Ph.D.
Assistant Professor of Psychology
Department of Social Science and History
Silver Lake College
2406 Alverno Drive
Manitowoc, WI 54220
Voice: (920) 686-6227
e-mail: [EMAIL PROTECTED]
_/_/_/_/_/ _/ _/_/_/_/
_/ _/ _/
_/_/_/_/_/ _/ _/
_/ _/ _/
_/_/_/_/_/ _/_/_/_/_/ _/_/_/_/_/
**********************************************************
http://www.sl.edu/socscience/Default.htm
