I am pleased to announce the latest release of the Real Statistics Resource Pack, Release 2.0. This release introduces multivariate statistics support, as described below. A complete description of the new capabilities is available by selecting the Multivariate Statistics tab from the home menu on the website.
The new release is available immediately for free download by clicking on the Download tab of the home menu on the website. Key new features are:
Hotelling’s T-square functions: Enables you to perform Hotelling’s T-square test for one sample, paired samples and independent samples (with equal covariance matrices and when covariance matrices are unequal). These tests are an extension of the t tests into the multivariate domain.
Box’s Test functions: Enables you to determine whether two or more covariance matrices are equivalent.
Mahalanobis distance squared function: Enables you to determine the distance between any data vector and the mean vector of a sample.
MANOVA functions: Enables you to perform multivariate analysis of variance. These are extensions of one factor ANOVA into the multivariate domain. The Wilks Lambda, Hotelling-Lawley Trace and Pillai-Bartlett Trace versions of the tests are supported. Two versions of Roy’s Largest Root are also calculated as well as the T, H and E SSCP matrices.
Extraction functions: Enables you to extract data from a multivariate sample for follow up analyses. The specific functions, ExtractRows, ExtractCol, ExtractCov, allow you to select only those sample elements in a particular group (i.e. independent variable) or dependent variable.
MANOVA data analysis tool: Automates the calculation of the MANOVA tests described above. In addition it outputs the covariance matrices for each group, the overall correlation matrix (to test for collinearity). It also identifies potential multivariate outliers (using the Mahalanobis distance squared) and automates Box’s Test to check for homogeneity of covariance matrices.
The data analysis tool also supports follow up tests using ANOVA on the samples for each independent variable as well as multivariate and univariate contrasts.
Multivariate Examples Workbook: In addition to the Real Statistics Examples Workbook, which contains all the examples found on this website prior to Release 2.0, you can now download for free a file with the new examples related to multivariate statistics. This file is called Real Statistics Mutivariate Examples and is available for download by clicking on the Download tab of the home menu on the website.
Upcoming releases: The new release does not yet include Factor Analysis. This capability will be available in Release 2.1 in a few days.
As usual I welcome you feedback regarding the site, future capabilities, bugs, questions, etc.
I hope that you find the new capabilities useful and that you continue to use the Real Statistics site and software.