Re: Factor Analysis Question

Richard Bell (rcbell@rubens.its.unimelb.edu.au)
Mon, 21 Jun 1999 08:32:36 +1000

Jonathan,

Negative factor loadings don't refer to the polarity of constructs by
themselves. They refer to the relationship with other constructs loading on
a factor. A negative loading means the construct has a reverse relationship
with the positive loading constructs. If you look at the construct
correlation matrix, there ought to be some negative correlations between
the constructs that relate to the factor you are considering. So if 'spends
money' has a negative loading there are probably other constructs (such as
'anal-retentive' or 'tight-arse') which have positive loadings but are
inversely related to 'spends money'

Richard

At 06:10 18/06/99 -0400, you wrote:
>Hello everyone. I am using the paper and pencil rep grid to get a better
>understanding of how high medium and low procrastinators describe their
>academic tasks. In order to run a factor analysis, I am getting
>participants to indicate with a tick, the more negative of the
>constructs and then to rate how that construct applies to each academic
>task. When I run the factor analysis, I (obviously)
>sometimes obtain negative factor loadings. When I ask my participants to
>label the grouping of constructs (i.e., to supply a label to each the
>cluster of constructs), I was wondering how I should tell them to
>interpret the negative factor loading. For example, one participant used
>"needs to be done" and "spend money" as the pair of constructs that
>emerged when we asked her how two of the three tasks were alike and one
>different. She then labelled 'spend money' as being the more negative of
>the two then rated that construct against all the academic tasks. In the
>factor analysis, 'spend money' loaded negatively on factor one. How would I
>tell her to interpret this construct? Would I tell her to think of what
>the opposite of spending money is or would I just scrap the 'spending
>money' construct and replace it with 'needs to be done' when showing her
>the list of words that loaded on factor 1? I hope I have not confused
>anyone. Thanks for you help in advance.
>
>-jl
>----------------------------------------------------------------------
>Jonathan Lee II Year M.A. Psychology
>Carleton University
>Email address: jmlee@chat.carleton.ca
> -Girlz an' Girlz They Love Me
> 'Cause I'm The Underweight
> Lover Jonnie Lee!-
>
>You pile up enough tomorrows, and you'll find you've collected a lot of
>empty yesterdays. http://www.carleton.ca/~tpychyl
>-----------------------------------------------------------------------
>
Richard C. Bell
Associate Professor
Department of Psychology
University of Melbourne
Parkville Vic 3052 Australia

ph: +61 3 9344 6364
fax: +61 3 9347 6618

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%