- Contents
- Learn about JMP
- Introduction to Multivariate Analysis
- Correlations and Multivariate Techniques
- Explore the Multidimensional Behavior of Variables
- Launch the Multivariate Platform
- The Multivariate Report
- Multivariate Platform Options
- Example of Item Reliability
- Computations and Statistical Details
- Principal Components
- Discriminant Analysis
- Partial Least Squares Models
- Develop Models Using Correlations between Ys and Xs
- Overview of the Partial Least Squares Platform
- Example of Partial Least Squares
- Launch the Partial Least Squares Platform
- Model Launch Control Panel
- Partial Least Squares Report
- Partial Least Squares Options
- Model Fit Options
- Statistical Details
- Hierarchical Cluster
- K Means Cluster
- Normal Mixtures
- Group Observations Using Probabilities
- Normal Mixtures Clustering Platform Overview
- Example of Normal Mixtures Clustering
- Launch the Normal Mixtures Clustering Platform
- Iterative Clustering Report
- Iterative Clustering Control Panel
- Normal Mixtures NCluster=<k> Report
- Robust Normal Mixtures
- Statistical Details for the Normal Mixtures Method
- Latent Class Analysis
- Group Observations of Categorical Variables
- Latent Class Analysis Platform Overview
- Example of Latent Class Analysis
- Launch the Latent Class Analysis Platform
- The Latent Class Analysis Report
- Latent Class Analysis Platform Options
- Additional Example: Plot Probabilities of Cluster Membership
- Statistical Details for the Latent Class Analysis Platform
- Cluster Variables
- Group Similar Variables into Representative Groups
- Cluster Variables Platform Overview
- Example of the Cluster Variables Platform
- Launch the Cluster Variables Platform
- The Cluster Variables Report
- Cluster Variables Platform Options
- Additional Examples of the Cluster Variables Platform
- Statistical Details for the Cluster Variables Platform
- Statistical Details
- References
- Index
Chapter 3
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
Figure 3.1 Example of a Multivariate Report
-
No Comment
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.