Dear John,
yes I know this paper. But let me add some notes:
a) Relaxing constraints (measurement errors/cross-loadings) in a completely
exploratory manner weakens the test of a model. Furthermore, relaxing
constraints can mask/avoid the misfit resulting from a too few numbers of
factors (that would otherwise be detected/corrected)
b) CFA does not *imply* non-existence of cross-loadings. When the
researchers thinks that a measure is
caused by two factors, theres no problem to incorporate these. Its the
data driven approach that is
problematic.
c) Cross loadings are no inherent attribute of personality measures. When a
researcher theorizes a given
number of factors with a specific meaning (which should be the case for the
Big 5 in the meantime),
nothing prevents him/her to develop measures that clearly measure each
factor (without cross loadings)
The cross-loadings are the consequence of the inductive history of the Big 5
(i.e., the lexical approach)
and the sometimes constructivistic view of those researchers on personality
(i.e., that a personality
"factor" is *constituted/defined* by a bunch of measures. According to this
view, the measures have to incorporate facets or at least some semantic
heterogeneity to "embrace" the "factor". However, such a view deviates from
a latent variable model (in which the factor is a single/uni-dimensional
entity that causes the measure.
Principally, a measure that measures two things at a time, is no good
measure (we don't want weighing
machines that measure our weight and the room temperature). Thus, we should
strive to develop clear measures instead of using the cross-loadings issue
to excuse the dramatic misfit of the
models ;) *polemic mode OFF*
I can really recommend the papers by the Dutch psychometrician Denny
Borsboom who really makes a good case (you can download all papers from
http://sites.google.com/site/borsboomdenny/dennyborsboom )
My favorites:
Borsboom, D., Mellenbergh, G. J., & van Heerden, J. (2004). The concept of
validity. Psychological Review, 111(4), 1061-1071.
Borsboom, D., Mellenbergh, G. J., & van Heerden, J. (2003). The theoretical
status of latent variables. Psychological Review, 110(2), 203-219.
Just my 2c
With warm regards,
Holger
Von: Organizational Behavior Division Listserv [mailto:
OB@AOMLISTS.PACE.EDU]
Im Auftrag von Kammeyer-Mueller,John Daniel
Gesendet: Freitag, 21. Januar 2011 16:31
An:
OB@AOMLISTS.PACE.EDU
Betreff: Re: [OB-LIST] AW: [OB-LIST] Which is the most frequently used Big 5
instrument?
Those interested in the factor structure of the FFM and its problems with
CFA might be interested in the following recent article by Herb Marsh and
colleagues, which argues the imposition of zero cross-loadings of items
(typical CFA practice) may be inappropriately restrictive for personality
scales. They propose an alternative structural equation model that does
support the existence of five relatively clear factors, consistent with
those found in traditional EFA:
Marsh, H. W., Lüdtke, O., Muthén, B., Asparouhov, T., Morin, A. J. S.,
Trautwein, U., & Nagengast, B. (2010). A new look at the big five factor
structure through exploratory structural equation modeling. Psychological
Assessment, 22, 471-491.
________________________________________
From: Organizational Behavior Division Listserv [
OB@AOMLISTS.PACE.EDU] On
Behalf Of Holger Steinmetz [
Holger.Steinmetz@WEB.DE]
Sent: Friday, January 21, 2011 3:11 AM
To:
OB@AOMLISTS.PACE.EDU
Subject: [OB-LIST] AW: [OB-LIST] Which is the most frequently used Big 5
instrument?
Hi folks,
just as a side note: Whenever I (or my students) have used Big 5 measures in
the last decade, NONE has ever survived a strict method (e.g., CFA).
Since the basis of validity is accordance to a proposed measurement
structure, I find it very strange that we rely on a personality model that
lacks a sound foundation. It seems that tradition beats evidence
. ;)
See the critisicm by
Borsboom, D. (2006). The attack of the psychometricians. Psychometrika,
71(3), 425-440.
And one additional note: Since the latent structure is so ambiguous (and
probably much more diverse than expected), the alphas
arent of much value (that means: alphas require essential tau-equivalence
that is strongly violated in this case).
Best,
Holger
___________________________________________________
Dr. Holger Steinmetz
University of Giessen
Faculty of Economics and Business Administration
Department of Human Resource Management, Small Business Enterprises,
and Entrepreneurship
Licher Str. 66
35394 Giessen
Phone: +49 641 99 22103
Fax: +49 641 99 22109
Holger.Steinmetz@psychol.uni-giessen.de
http://wiwi.uni-giessen.de/home/personal
___________________________________________________
Von: Organizational Behavior Division Listserv [mailto:
OB@AOMLISTS.PACE.EDU]
Im Auftrag von Taya Cohen
Gesendet: Donnerstag, 20. Januar 2011 18:48
An:
OB@AOMLISTS.PACE.EDU
Betreff: Re: [OB-LIST] Which is the most frequently used Big 5 instrument?
Dear Stefan,
You might consider the HEXACO Personality Inventory. In addition to the
traditional Big 5 factors, the HEXACO also assesses a 6th
dimension--Honesty-Humility--that predicts ethical decision-making and
behavior. Kibeom Lee & Michael Ashton maintain a website where you can
download the measure and relevant articles:
http://hexaco.org.
Taya
----------------------------------------------------------------------------
-----------------------------------
Taya R. Cohen
Assistant Professor of Organizational Behavior & Theory
Tepper School of Business | Carnegie Mellon University
Email:
tcohen@cmu.edu
Office Phone: (412) 268-6677
Web:
http://taya.cohen.socialpsychology.org/
----------------------------------------------------------------------------
-----------------------------------
On Wed, Jan 19, 2011 at 7:44 AM, Stefan Volk <
stefan.volk@uni-tuebingen.de>
wrote:
Dear all,
in recent years we have seen a sharp increase in the number of available Big
5 instruments, from very short (5 items, e.g. Woods and Hampson, 2005) to
very long (240 items). When choosing an instrument there is of course always
the trade-off between brevity and depth. I was wondering which Big 5
instrument has in your view the best cost/benefit ratio, or, put it
differently, which is the most frequently used Big 5 instrument out there?
I would highly appreciate if you could share your experience: Which
instrument have you used most frequently and what are the pros and cons in
your view? The answer to my question depends of course on the purpose of the
study, it would therefore be great if you could also briefly indicate in
which research context you are using the Big 5. I will of course prepare a
summary document with all answers for the list and we may also prepare a
short research note for the wider community.
I have frequently used the TIPI (Gosling et al., 2003) in combination with
economic experiments. Pro: short, widely used, acceptable psychometric
properties. Con: provides no facet-level scores, obviously low Cronbachs
alphas and inter-item correlations (in particular for Agreeableness), which
is frequently picked up by reviewers.
Many thanks in advance,
Stefan
--
Dr. Stefan Volk
Assistant Professor
Tübingen University Department of International Business
Melanchthonstr. 30
72074 Tübingen
Germany
Phone +49 7071 29 74050
Fax +49 7071 29 5534