Multivariate Methods • ­Correlations and Multivariate Techniques
Previous
 • 
Next
­Correlations and Multivariate Techniques
Explore the Multidimensional Behavior of Variables
Use the Multivariate platform to explore how many variables relate to each other. The word multivariate simply means involving many variables instead of one (univariate) or two (bivariate). From the Multivariate report, you can:
•
summarize the strength of the linear relationships between each pair of response variables using the Correlations table
•
identify dependencies, outliers, and clusters using the Scatterplot Matrix
•
use other techniques to examine multiple variables, such as partial, inverse, and pairwise correlations, covariance matrices, principal components, and more
Example of a Multivariate Report
JMP Learning Library – Correlation and Regression
Watch a brief video on how to obtain correlations between quantitative measures, including multivariate correlations.
Contents 
Launch the Multivariate Platform
Estimation Methods
The Multivariate Report
Multivariate Platform Options
Nonparametric Correlations
Scatterplot Matrix
Outlier Analysis
Item Reliability
Impute Missing Data
Example of Item Reliability
Computations and Statistical Details
Estimation Methods
Pearson Product-Moment Correlation
Nonparametric Measures of Association
Inverse Correlation Matrix
Distance Measures
Cronbach’s α