Meaning of Misspattern in the ods output ModelAnova

[Pierre_Ars]

Hello,

I am running a proc anova and in the output Modelanova I see a variable Misspattern which seems to take 3 different values (in my case) : 1,2,3.

For example : 

MissPattern	Dependent	HypothesisType	Source	DF	SS	MS	FValue	ProbF
1	m_chargetot	6	CD_LOB_NOMAD	1	22733363953	22733363953	6,845565506	0,00904096

Can someone explain how to interpret those values (here 1)?

Thanks in advance !

[PaigeMiller]

There are three different missing patterns in your X-data. For example, when MISSPATTERN = 1 then X1 is missing, when MISSPATTERN = 2 then X2 and X3 are missing; and when MISSPATTERN = 3 then X1 and X3 are missing.

In the presence of missing X variables, you should not be using PROC ANOVA. The documentation states clearly:

Use PROC ANOVA for the analysis of balanced data only, with the following exceptions: one-way analysis of variance, Latin square designs, certain partially balanced incomplete block designs, completely nested (hierarchical) designs, and designs with cell frequencies that are proportional to each other and are also proportional to the background population. These exceptions have designs in which the factors are all orthogonal to each other.

[Pierre_Ars]

Thank you !

But where can I find the interpretation of MissPattern (=1,2,3,...)  in the documentation ? How did you know that Misspattern = 2 implies X2 and X3 are missing ? I made a lot of tests and I always got a value of 1,2 or 3 for this variable (never 0 for example)...

Thank you in advance !

[PaigeMiller · Accepted solution]

The interpretation of MISSPATTERN is found by LOOKING AT your data with your own eyes and seeing what the pattern of missing is that exists in the data set.

How did you know that Misspattern = 2 implies X2 and X3 are missing ?

I did not say that MISSPATTERN = 2 implies that X2 and X3 are missing. I was giving an example, which began with the words "For Example", so it could be other variables that are missing — which you can figure out yourself by LOOKING AT your data with your own eyes.

[Pierre_Ars]

So thank you @PaigeMiller , your point allows me to solve my problem !