Tell me more ×
Cross Validated is a question and answer site for statisticians, data analysts, data miners and data visualization experts. It's 100% free, no registration required.
up vote 1 down vote favorite

Hello and thanks in advance for your help!

I have conducted and EFA for a scale I made that includes 13 items. Of these items, 2 are recoded. Looking at the rotated component matrix I see that this scale has 2 factors, one with 11 items and one with the 2 recoded items. What does this mean and how do I decribe this? Does anyone know of a journal article or citation that deals with this sort of issue.

share|improve this question
1  
Can you state what you mean by "recoded". EG, if it were just reverse scored, it would still load on the same factor, but the sign of the loading would flip. It may be just a coincidence that the only 2 recoded items are different. – gung Dec 5 '12 at 13:53
Yes, reverse scored and I know that it shouldn't matter but could it just be a coincidence that they happen to be reverse scored and are another factor? – Noah41 Dec 5 '12 at 13:55

1 Answer

up vote 3 down vote

As noted in gung's comment, mathematically speaking reverse scoring does not make any difference and it might very well be a coincidence that the recoded items end up together.

It is however also conceivable that a tendency to agree with all items independently of their content (“acquiescence bias”) would attenuate the correlations between positively and negatively worded items and consequently lead to the presence of a spurious factor.

I haven't read it but this paper might be relevant to your situation: Schmitt, N. & Stults, D.M. (1985). Factors defined by negatively keyed items: The result of careless respondents? Applied Psychological Measurement, 9, 367-373.

share|improve this answer
Thank you! Your help is greatly appreciated – Noah41 Dec 7 '12 at 10:57
Excellent points! Also, @Noah41 - since you say you obtained a "rotated component matrix," I'm betting you used principal components analysis when exploratory factor analysis proper would have been much better. A very common mistake. See stats.stackexchange.com/questions/1576/… – rolando2 Jan 5 at 15:36

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.