Re: factor Analysis
It's not so simple. You have to do matrix-inversion for that. If your statistical program is able to spit out factor scores, you just take these as your coordinates. For each of your objects you get values in each factor, which you can use as coordinates in the factorspace. Regards - Gottfried. Huxley schrieb: Thank you for explanation. Bu my question was unclear therefore let me ask again. I invented an exapmle. I have 10 questions in a questionnaire. These questions are my 10 variables. A consumers fill this questionnaire for each 15 products e.g cars. Because 10 variables (X1, X2, ...,X10) are correlated with each other I use factor analysis and (for convinence I ordered it) I get Factor1: X1,X2,X3,X4,X5,X6,X7 Factor2: X8,X9,X10 I can e.g put X1 into 2-D space, because I know that X1= -1*F1+ (-1*F2). It means that X1 has co-ordinates X1=(-1,-1). It's simple. But I'm not interested in positioning X1. For me it's important where there are products (cars) in 2-D space. Therefore my question is how to do it. I heard (but I do not know) that using e.g variable X1,...X10 mean and factor loadings I can do it i.e. for car1: I multiple factor loadings and variables mean (suitable) and I get this position Could you help me verify this? I would be very appreciate Regards Huxley = Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at http://jse.stat.ncsu.edu/ =
Re: factor Analysis
Huxley schrieb: Uzytkownik Gottfried Helms [EMAIL PROTECTED] napisal w wiadomosci [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... It's not so simple. You have to do matrix-inversion for that. Not simple? I heard that taking suitable factor loadings and every variable mean I can obtain this space. e.g. (I do not know is it true) Let mean for car1 and questions 10 (variables): mean X1=1 mean X2=2 .. mean X10=10 I have 2 factor score. factor loadins (aij) I have, therefore for first factor score, co-odrinate for car1 is F1(for car1)=1*a(1,1)+2*a(2,1)+3*a(3,1)+...+10*a(10,1) is it true? Huxley Loadings of factor f1,f2 for items x1,x2,x3,x4... f1f2 x1 0.4 0.6 x2 0.3 0.9 x3 0.2 -0.1 x4 -0.8 -0.4 ... Call this loadingsmatrix A, your correlation-matrix R That means, that A*A' = R Call your empical datamatrix (x1,x2,x3,...) X Call the unknow factorscores SC Then it is assumed that A*SC = X Then you must find inv(A) to be able to find SC: inv(A)*A*SC = inv(A) *X SC = inv(A)*X If the shape of A is not square and/or the rank is lower then its dimension, then you have to find a workaround to compute the general_inverse of A. I don't find it so simple ;-) Gottfried. = Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at http://jse.stat.ncsu.edu/ =
tricky explanation problem with chi-square on multinomial
of the squared deviation - up to a local maximum. My difficulties are, to make this clear in simple words; best in such simple words, as I used, when I explained the rationale of chi-square and significance... Ok, maybe, it's more a subject for news://sci.stat,edu , I guess. Thanks again for your input - Gottfried Helms. -- 0 0 0 0 0 0 0 0 0 0 0 (need more text for being able to send. Thx Netscape! 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 = Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at http://jse.stat.ncsu.edu/ =
Re: tricky explanation problem with chi-square on multinomial
Hi Jos, got your msg. Thanks! You might consider the distribution of chi-square(df) / (df), which as far as I know has not been given a name; this distribution would be concentrated around expectation 1 with variance 2/(df). Seems to be reasonable. Like using Cramer's V instead of Chi-square. The actual problem is that of how to translate this to students, who are used to: the farer away from expectation (i.e. uniformity) the more unlikely is the outcome. Or opposite: the expected is the most likely. If the uniformity is not the most likely, why does it still engaged as the expected, from where we calculate deviations? They have to learn a different slogan, i'm afraid... Gottfried. --- Jos Jansen schrieb: (...) I hope this will clear up the matter a little. Jos Jansen = Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at http://jse.stat.ncsu.edu/ =
Re: tricky explanation ... /pls. excuse
Sorry, didn't realize, the cited comment was only private mail. Just assumed, it were NG pm. Pls. excuse Gottfried. = Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at http://jse.stat.ncsu.edu/ =
Re: Texts: Factor Analysis
[EMAIL PROTECTED] wrote: What are your favorite book(s) on factor analysis? What do you think of R. Gorsuch's book? My favorite is Stan Mulaik "The foundations of factor analysis". It is comprehensive and still straightforward from the introduction to all covered themes. I have tried different others, but none was like that. Not being educated mathematician I felt I got most that I needed with a good insight of the principles. One similar is from Dirk Revenstorf, but I doubt it is available in english. Gottfried Helms. -- - Gottfried Helms Soz.Päd./Soz.Arb. FB04 // FG Prevention Rehabilitation at University D-34109 Kassel Moenchebergstr. 19 B email: mailto:[EMAIL PROTECTED] www: http://www.uni-kassel.de/~helms === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===