Re: Compare matrices using Chi square for Exploratory

*To*: mathgroup at smc.vnet.net*Subject*: [mg114870] Re: Compare matrices using Chi square for Exploratory*From*: Stuart Nettleton <Stuart.Nettleton at uts.edu.au>*Date*: Tue, 21 Dec 2010 00:18:00 -0500 (EST)*References*: <201012200539.AAA22748@smc.vnet.net>

A small addendum to my previous question with the issues that gave rise to my question about a simple approach to a whole-of-matrix comparison between original and recreated data. I tracked down R's chi-squared comparison to Lawley & Maxwell (1962 & 1971). Their Bartlett test suffers from a few disadvantages such as requiring the inverse of what I have discovered to be an ill-conditioned matrix or working from correlation matrices rather than original data. My preference is to use Mathematica's elegant automatic Hypothesis tests in VarianceEquivalenceTest. At present this function compares vectors of data, rather than matrices. Mapping across each pair of original and recreated variables will lead to as many tests results as variables. This would be quite cumbersome and somewhat analogous to investigating the percentage of differences between correlation matrices that are greater than say 0.05, which is interesting but not a test upon which to accept or reject the number of factors. Thus my question. Many thanks to those who may be looking at this question. BTW For those using a Version prior to 8.0, the first line of the package would need amendment to BeginPackage["zigzagefa`",{"MultivariateStatistics`"}] UTS CRICOS Provider Code: 00099F DISCLAIMER: This email message and any accompanying attachments may contain confidential information. If you are not the intended recipient, do not read, use, disseminate, distribute or copy this message or attachments. If you have received this message in error, please notify the sender immediately and delete this message. Any views expressed in this message are those of the individual sender, except where the sender expressly, and with authority, states them to be the views of the University of Technology Sydney. Before opening any attachments, please check them for viruses and defects. Think. Green. Do. Please consider the environment before printing this email.

**References**:**Compare matrices using Chi square for Exploratory Factor Analysis***From:*Stuart Nettleton <Stuart.Nettleton@uts.edu.au>