Is there a procedure to check whether missing data are missing completely at random or missing at random?

[k_ecu03]

I was just wondering if there was a procedure to check whether missing data are MCAR or MAR?

[mohamed_zaki]

Not sure, but you can:

1-Use the Little's MCAR Test Macro implemented in http://www.appliedmissingdata.com/macro-programs.html

2-Testing your missing data against known missing patterns as suggested in Knowledge (of your missing data) is power: handling missing values in your SAS® dataset using the %missingPattern as described in Studying missing data patterns using a SAS macro.

[Ksharp]

My thought is checking their frequency, and make a hypothesis test H0: mean of freq =xx .

[k_ecu03]

Thanks to all. My sample does not meet the normality assumptions, so I used the Whitney-Mann to test between the complete and missing scores. I appreciate the replies.

[mohamed_zaki]

then

please mark the question as answered.