Re: Question about negative values in rotated matrix in factor analysis?

The "index" variable that you are creating is also called a summative scale. Scores are conventionally created by "reflecting" items that load negatively . As z-scores you can just multiply by -1 (i.e., switch the signs. Then simply sum the items that load 1) above some criterion absolute value, e.g., .4 and 2)load cleanly (i.e., don't load much on another factor) and 3) make sense in terms of a construct.

Also be careful not to retain too many factors. the Kaiser criterion is used only to give a maximum for extraction, It is rarely, the number to retain in an interpreted solution.

Art Kendall
Social Research Consultants



jstankov@xxxxxxxxxxx wrote:
I have a question about negative values in factor analysis. I am
using factor analysis (principal components with varimax rotation) to
create some index variables for a future cluster analysis. I z-scored
about 60 variables and then ran the factor analysis. However, in the
rotated matrix, I have a few high negative values (eg, -.75) that are
in the same column as high positives (eg, 0.80). Obviously, the two
variables are negatively related and that makes sense. For example, I
am looking at colleges in the US and Admissions Yield is negative and
Tuition is positive. My question is: Say the rotated component
has 5 high positives and 1 high negative. Can I add the 6 z-scored
variables together and divided by 6 to create an index? Or, do I
leave out the negative variable? Or, do I reverse the negative
variable in some way -- if so, how?

Thanks in advance,
Joe

.

Relevant Pages

  • Re: Question about negative values in rotated matrix in factor analysis?
    ... As z-scores you can just multiply by -1 (i.e., ... using factor analysis (principal components with varimax rotation) to ... rotated matrix, I have a few high negative values that are ... in the same column as high positives ...
    (sci.stat.math)
  • Question about negative values in rotated matrix in factor analysis?
    ... using factor analysis (principal components with varimax rotation) to ... create some index variables for a future cluster analysis. ... rotated matrix, I have a few high negative values that are ... in the same column as high positives ...
    (sci.stat.edu)
  • Question about negative values in rotated matrix in factor analysis?
    ... using factor analysis (principal components with varimax rotation) to ... create some index variables for a future cluster analysis. ... rotated matrix, I have a few high negative values that are ... in the same column as high positives ...
    (sci.stat.math)
  • Re: Principal Component Analysis- Do I need to scale (i.e. normalize) my variables?
    ... When you tell the software to use the covariance matrix in any type of factor analysis (principal components, principal factors, alpha, image, etc) the scales are important. ... where you tell the software to use the correlation matrix the variables are implicitly standardized. ... I don't know what the substantive meaning of x's would be if you used the absolute value of the z-scores aka standardized variables. ... I need to perform PCA on 20 or so variables (ex. ...
    (sci.stat.math)
Loading