Classify Missing Data as MCAR, MAR, or MNAR

Classify missing data as missing completely at random (MCAR), missing at random (MAR), or missing not at random (MNAR). This step is required before handling missing data (e.g. mean imputation) so that bias is not introduced. See Little (1988) for the statistical rationale for the methods used.

missr

The goal of missr is to help you classify missing data as MCAR, MAR, or MNAR. It does this by providing:

• Statistical tests for MCAR and MAR:
  • mcar()
  • mar()
• Process of elimination for MNAR:
  • mnar()

Installation

You can install missr from CRAN:

install.packages("missr")

Or you can install the development version on github using remotes:

# install.packages("remotes")
remotes::install_github("NoahHellen/missr")

Examples

• Testing for MCAR.
  • Null hypothesis is that data is MCAR; if the p-value is not significant, there is evidence the data is MCAR.

mcar(data)

• Testing for MAR.
  • Each null hypothesis is that data is not MAR; if each p-value is significant, there is evidence the data is MAR.

mar(data)

• Testing for MNAR.
  • No new tests, simply a process of elimination.

mnar(data)

License

missr has an MIT license, as found in the LICENSE file.