Articles in this section

How to Create a Scatterplot Matrix

  • Updated
Follow

This article describes how to create a correlation scatterplot matrix from a series of variables. The correlation scatterplot matrix shows a matrix where the lower triangle displays scatterplots of variable pairs, the diagonal displays variable histograms and the upper triangle displays correlations of variable pairs. Also known as a SPLOM.

mceclip0.png

Requirements

  • A series of categorical or numeric variables to use as correlation inputs. 
  • See Variable Sets for more info on how to set the measurement scale and values for analysis.

 

Method

  1. From the toolbar, select Anything > Advanced Analysis > Correlation > Scatterplot Matrix. 
  2. From the object inspector, select variables from the Variables to plot drop-down or drag them from the Data Sets tree into the Variables to plot box. 
  3. Click Calculate.
  4. OPTIONAL: Variable names - toggle this on to show variable names in the matrix instead of variable labels.

  5. Missing data: Method for dealing with missing data.

    • Error if missing data - An error is returned if any of the data used in the analysis contains missing values.
    • Exclude cases with missing data - The analysis is conducted using cases with no missing data. For example, if there are three variables, x, y, and z, and the total sample size is 10, but 5 cases have no data for z, only the 5 cases with the complete data are used in the analysis. This is also known as casewise deletion and the complete-case method. It is the default approach in Displayr.
    • Use partial data - The analysis is conducted using all the data for each case. For example, in K-Means Cluster Analysis, if there are nine variables in the analysis, and a case only has data for six, then the case is assigned to the most similar cluster based on the data for the six variables.
  6. Fitted line:
    • None - No fitted line is shown in each scatterplot.
    • Straight - A line of best fit (from linear regression) is shown in each scatterplot.
    • LOESS - A line generated using fitted values from locally weighted polynomial regression is shown in each scatterplot.
  7. Scatter point modifications:
    • None - No modifications.
    • Jitter - Add noise to each point so that areas with closely positioned points appear denser.
    • Enlarge points with multiple observations - Multiple points with identical positions will appear as an enlarged point. This is the default.

OPTIONAL: Further visual modifications can be made via the Chart tab in the object inspector.

This tab contain options for formatting the text and graphical elements in the chart. It is organized into groups for CORRELATION (panels in the upper-right triangle), SCATTERPLOTS (panels in the lower-left triangle) and HISTOGRAMS (panels along the diagonal).

 

Next

How to Create a Correlation Matrix

How to Conduct a Test of Correlation Between Two Variables

How to Create a Table of Coefficients from a Correlation Matrix

 

Related articles