I am using exploratory factor analysis on a rather large survey

instrument. Several of the variables are loaded heavily (>|.3|) on

different factors. Should such variables be tossed out, or is there a

better way of dealing with the problem?

--

Adam McKee

___________________________________

work phone: 266-5460

home phone: 271-8247

I think that this feature is inevitable.

The clearest results you will become after oblique rotation.

--

OUTLIER, Consultants in Statistics

www.outlier.be

> I am using exploratory factor analysis on a rather large survey

> instrument. Several of the variables are loaded heavily (>|.3|) on

> different factors. Should such variables be tossed out, or is there a

> better way of dealing with the problem?

> --

> Adam McKee

> ___________________________________

> work phone: 266-5460

> home phone: 271-8247

> instrument. Several of the variables are loaded heavily (>|.3|) on

> different factors. Should such variables be tossed out, or is there a

> better way of dealing with the problem?

> --

> Adam McKee

> ___________________________________

> work phone: 266-5460

> home phone: 271-8247

On Wed, 27 Oct 1999 10:07:02 -0500, Adam McKee

> I am using exploratory factor analysis on a rather large survey

> instrument. Several of the variables are loaded heavily (>|.3|) on

> different factors. Should such variables be tossed out, or is there a

> better way of dealing with the problem?

distinguish the factor structure. If that were the case, the main

solution would be,

Get a much larger sample or use a lot fewer variables.

--

http://www.pitt.edu/~wpilib/index.html

I would not consider |.3| as a high factor loading. It's reasonable to consider

factor loadings above |.6| as high. Also, make sure that the factor loadings

you look at are the ones displayed after rotation.

Hi

> I am using exploratory factor analysis on a rather large survey

> instrument. Several of the variables are loaded heavily (>|.3|) on

> different factors. Should such variables be tossed out, or is there a

> better way of dealing with the problem?

good reason to think that each variable (whether single item or a

composite) is entirely "pure" (i.e., measures a single

dimension). So I would take such results (presuming the loadings

are large enough, as noted by another respondent) as a hint at

the multidimensional nature of the variable. You want to ask

yourself whether it makes sense that specific variables load on

different factors, perhaps by examining variables that are more

pure indicators of the factor.

Best wishes

Jim

===========================================================================

James M. Clark (204) 786-9757

Department of Psychology (204) 774-4134 Fax

University of Winnipeg 4L05D

CANADA http://www.uwinnipeg.ca/~clark

===========================================================================

Hello

We have a data set which includes, for each of 96 cities, 4 estimates

of the number of drug injectors in that city. Each of these estimates

contains error.

We ran a factor analysis to determine their commonality.

We now wish to use the results of this analysis to come up with a

better estimate of the number of drug injectors in each city.

It seems intuitively reasonable to multiply the standardized scoring

coefficients by the estimates, add these together, and then divide by

the sum of the standardized scoring coefficients. I've even seen this

done. But I haven't seen a good proof that this is correct (it may not

be correct!)

Any advice on how to proceed will be appreciated.

Thanks in advance

Peter L. Flom, PhD

Assistant Director, Statistics and Data Analysis Core

Center for Drug Use and HIV Research

National Development and Research Institutes

71 W. 23rd St

New York, NY 10010

(212) 845-4485 (voice)

(917) 438-0894 (fax)

2. Bruno Newsletter - how does it rate?

5. Oracle BAP

7. Need NaturallySpeaking Preferred & Dragon NaturallySpeaking Legal Suite (New or Used)

8. Please ignore/excuse exploratory factor analysis question

10. factor analysis: calculate factor scores?

11. Calculating factor scores for new dataset after factor analysis

12. Three questions about factor analysis (scree, Bartlett, cumulative over 1.0)

5 post • Page:**1** of **1**