Enter your data for Principal Components Analysis

Stat > Multivariate > Principal Components

Specify the data for your analysis, enter the number of components to calculate, and specify the type of matrix.

In This Topic

Enter your data
Number of components to compute
Type of Matrix

Enter your data

In Variables, specify the columns of data that you want to analyze. You must have two or more columns of numeric data, with each column representing a different measurement. If a missing value exists in any column, Minitab ignores the entire row. Minitab excludes missing values from the calculation of the correlation or covariance matrix.

In this worksheet, each column contains measurements for a different type of information on a loan application.

C1	C2	C3	C4	C5	C6	C7	C8
Income	Education	Age	Residence	Employ	Savings	Debt	Credit cards
50000	16	28	2	2	5000	1200	2
72000	18	35	10	8	12000	5400	4
61000	18	36	6	5	15000	1000	2
88000	20	35	4	4	980	1100	4
91100	18	38	8	9	20000	0	1
45100	14	41	15	14	3900	22000	4

Number of components to compute

Enter the number of principal components that you want Minitab to calculate. If you have a large number of variables, you may want to specify a smaller number of components to reduce the amount of output. If you do not know how many components to enter, you can leave this field blank. If you do not specify a number, Minitab calculates the maximum number of components, which equals the number of variables. You can then use the output to determine how many components explain most of the variation in the original variables.

Type of Matrix

Select the type of matrix to use to calculate the principal components.

Correlation: Use when your variables have different scales and you want to weight all the variables equally. For example, if some of the variables use a scale from 1-5 and others use a scale from 1-10, use the correlation matrix to standardize the scales.
Covariance: Use when your variables use the same scale, or when your variables have different scales but you want to give more emphasis to variables with higher variances.

For example, suppose you count different species of organisms at several sample sites. If you select the covariance matrix, the more common species will show higher variances and be given more emphasis. Very rare species will not affect the analysis as much. If you select a correlation matrix, all species are weighted equally. Therefore, very rare species may contribute significantly to the analysis results. Therefore, the decision depends on the objective of your study.