Dear list,

I have a student that has Rorschach protocols for 3 different groups of priests. He initially correlated stress scores from a scale with various measures of the Rorschach. He did this for each group of priests. I recommended that he use regression and dummy code the three groups of priests. He wants to compare each group to the other on the relation between the IV and DV. What would be the best way to accomplish this. Suggestions appreciated.

Martin Sherman

Does anyone know if Multivariate Regression (using multiple DVs and multiple IVs; I believe you can do this in SAS) is mathematically the same as doing an SPSS GLM Multivariate Analysis of Covariance (using multiple DVs and multiple covariates)?

It seems very similar, except that you don't get a regression output in GLM.

Thanks
Matt

Matthew Pirritano, Ph.D.
National Science Foundation Post-Doctoral Fellow
College of Education
Department of Individual, Family & Community Education
MSC05 3040
1 University of New Mexico
Albuquerque, NM 87131-0001
Telephone (505)277-7115
FAX (505)277-8361

Dear list, Is it possible to perform a spectral map analysis with SPSS? If not, some one knows a free program to do so. Many thanks, /Christian

I believe, but am not sure, that spectral map analysis is the same as correspondence analysis (dual scaling). It is certainly similar in that both spectral map analysis and correspondence analysis project variables and cases in a deduced dimensional space.

Every so often a procedure comes around that is the same but goes by different names. (e.g., correspondence analysis from Europe & dual scaling from Canada, and possibly a form of multidimensional scaling from the US). If you do no get closure from the responses on this list, I suggest you describe your data and problem on the Classification Society list. Mention that you would to use SPSS if possible. http://www.classification-society.org/csna/lists.html#class-l Art Art@DrKendall.org Social Research Consultants University Park, MD USA (Inside the Washington, DC beltway.) (301) 864-5570

Good Afternoon Friends, I have a survey composed of 10 yes/no items. I want to compare Yes responses to Item 5 with Yes responses to Item 9. I am interested in testing if there is a significant difference in Yes responses between Item 5 and Item 9. Is the McNemar appropriate? I remember vaguely that someone posted a macro for: test of significant difference between dependent proportions? I failed to find it at Raynald Levesque's page Raynald@spsstools.net .

Dear Experts, I am developing a measure on creative climate. I have one dimension that resulted in a Cronbach Alpha of .70. However, when I examine the loading of its items (n = 5) via Alpha Factor Analysis, the dimension is unstable, namely, its items do not load to form a factor. In fact, not one item representing this dimension loaded with any other-own item. How can I explain this to a lay person? Thanks. John

孩子咋了？

thank you  very much, it~s someting good to me