Plots

You can plot data and model results. At present, this section gives only a very rough overview.

Overview

The basic PLOT command

Requesting plots is rather simple, as you do not have many choices at this stage (choices will come later!). There are three (3) basic plot commands. The first one goes like this:

PLOT:
      TYPE IS PLOT1;

and you may add PLOT2 and/or PLOT3 or select only one of these three options. Actually, each of these commands typically requests several plots which can be viewed later on.

Basically, what you will get with the different options is the following:

  • PLOT1 will produce descriptive plots of the sample data. The specific type of plot created will depend on the type of data.
  • PLOT2 will display estimated means, probabilities and the like.
  • PLOT3 will graph estimated values for individual cases.

As a special feature of PLOT3, you may request graphs of outlier statistics. For this, you will need an extra keyword together with PLOT3 as follows:

PLOT:
      TYPE IS PLOT3;
      OUTLIERS ARE COOKS;

Instead of, or in addition to, COOKS the following statistics are available:

  • MAHALANOBIS: Mahalanobis distance (available only for continuous variables and only for some estimation methds)
  • LOGLIKELIHOOD: the contribution of each case to the log-likelihood (available only with ML estimation)
  • INFLUENCE: a loglikelihood distance influence measure, also called "likelihood displacement for maximum likelihood estimators". An analogous fit measure is available for the weighted least squares estimators.

Viewing and changing plots

To view a plot, choose GRAPH and than VIEW GRAPH from the from Mplus menu. From the output window, you may also click on the red V button on the toolbar. You now can choose from the plots that have been created. There are some, but not many, possibilities of plot manipulation.

Exporting plots

You can export a plot you are viewing by choosing GRAPH – EXPORT GRAPH from the menu. Plots can be exported in DIB, EMF or JPEG format.

© W. Ludwig-Mayerhofer, Mplus Guide | Last update: 15 Mar 2010