On Sun, 20 Jun 2004, PD Dr. Hans-Christian Waldmann wrote: > A client to our university stats consulting has produced a > questionaire with over 250 items that he considers to be > associated to 12 scales (with 6-30 items for a scale). There > is consent that the number of items is far too high in a > clinical context, and has asked us to reduce the item count > in a "pragmatic" way. Hm. He objects to the idea of re- > evaluating content validity ("they're all reasonable"). > Nevertheless, I've been asked to cut 30% of the items while > retaining the 12 scales.
You do not say whether you have pilot info on responses to all items, or how many subjects you have, if so. You might want to consider not only the homogeneity of the items (i.e., convergent validity), which your criteria primarily address, but also the discriminant validity (i.e., whether items correlate with other scales). One approach, given enough subjects, would be factor analysis and select items that load purely on the desired factors (i.e., loadings on only one scale). At least, items should correlate more highly with scale they are assigned to than to other scales.
Another consideration in scale creation (see Doug Jackson's work on personality, vocational, and clinical assessment) is correlation with various measures of response bias (e.g., Paulhus or other measures, desirability). It might be good, for example, if half the items on each scale were positively worded and half negatively worded.
Without knowing the domain, 12 seems like a lot of factors to try and get independent measures of, so results are likely to be pretty messy.
Best wishes Jim
============================================================================ James M. Clark (204) 786-9757 Department of Psychology (204) 774-4134 Fax University of Winnipeg 4L05D Winnipeg, Manitoba R3B 2E9 firstname.lastname@example.org CANADA http://www.uwinnipeg.ca/~clark ============================================================================