Cruise Scientific                Visual Statistics Studio                Visual Statistics Studio Guide

Data Prototypes

The Visual Statistics Studio designs, textbooks, guides, and manuals are conceptualized around the method of smallest possible data sets. The key methods of the general linear model are as follows

Method Minimum Set of Variables Minimum n Limiting Conditions Data Prototypes
Descriptive Statistics X 2   Data, Univariate Prototypes
Correlation X Y 3 For n equal to 2 r is either -1 or +1 Data, Bivariate Prototypes
Regression X1 Y 3 For n equal to 2 R equals 1 Data, Bivariate Prototypes
Multiple Regression X1 X2 Y 4 For n equal to number of variables R equals 1 Data, Multivariate Prototypes
Canonical Analysis X1 X2 Y1 Y2 5 For n equal to number of variables rho equals 1 Data, Multivariate Prototypes

Thus the minimum n for data sets of the general linear model is 5 with the basic data prototype X [1 2 3 4 5] and variations of the numbers 1 2 3 4 5 as [3 1 2 4 5], [2 3 1 5 4], etc. All basic data prototypes have mean equal to 3 and true variance equal to 2. Data prototypes for specific methods of the general linear model are located under the Designs menu and on the (Project, Open Project Files) menu. Data for distributions and function are stored under the Arguments menu.