In missing-data analysis, Little's test (1988, Journal of the American Statistical Association 83: 1198–1202) is useful for testing the assumption of missing completely at random for multivariate, partially observed quantitative data. I introduce the mcartest command, which implements Little's missing completely at random test and its extension for testing the covariate-dependent missingness. The command also includes an option to perform the likelihood-ratio test with adjustment for unequal variances. I illustrate the use of mcartest through an example and evaluate the finite-sample performance of these tests in simulation studies.
FitzmauriceG., DavidianM., VerbekeG., and MolenberghsG.2009. Handbooks of Modern Statistical Methods: Longitudinal Data Analysis.Boca Raton, FL: Chapman & Hall/CRC.
2.
KimK. H., and BentlerP. M.2002. Tests of homogeneity of means and covariance matrices for multivariate incomplete data. Psychometrika67: 609–623.
3.
LittleR. J. A.1988. A test of missing completely at random for multivariate data with missing values. Journal of the American Statistical Association83: 1198–1202.
4.
LittleR. J. A.1995. Modeling the drop-out mechanism in repeated-measures studies. Journal of the American Statistical Association90: 1112–1121.
5.
LittleR. J. A., and RubinD. B.2002. Statistical Analysis with Missing Data. 2nd ed. Hoboken, NJ: Wiley.
6.
RubinD. B.1976. Inference and missing data. Biometrika63: 581–592.
