How to Check Missing Data Using Little's MCAR Test

This article describes how to tests the null hypothesis that missing data is Missing Completely At Random (MCAR). A p.value of less than 0.05 is usually interpreted as being that the missing data is not MCAR (ie, is either Missing At Random or non-ignorable). See What are the Different Types of Missing Data? for more information about MCAR and other types of missing data.

Requirements

Please note these steps require a Displayr license.

A Displayr document with a data set.

Method

1. Go to + > Data > Missing Data > Little's MCAR Test.
2. Object inspector > Data > Variables select the variables you want to analyze and click Calculate to run the function (or drag them from the Data Sources into the Variables section).

OPTIONAL: You can adjust the following settings to modify the calculation and output:

• Output
  • Summary - Shows a nicely formatted table of the test results (default).
  • R - The original text-based output from the LittleMCAR function.
• Variable names - Display Variable Names in the output, instead of Variable Labels.
• More decimal places - Display numeric values with 8 decimal places.
• Filter - The data is automatically filtered using any filters prior to estimating the model.

Technical details:

This test starts by using the EM algorithm to estimate the means and covariances. These estimates are approximate, and, consequently, the test is also approximate (it may get slightly different results on different computers).

This test does not take weights into account.

Acknowledgements:

Uses the LittleMCAR function from the R Package BaylorEdPsych.

Little, R. J. A. (1988). A test of missing completely at random for multivariate data with missing values. Journal of the American Statistical Association, 83(404), 1198–1202.