Variables Missing Pattern

For a set of variables X1 X2,, Xi, this function is to report on their joint missing pattern.

The information on joint missing pattern is helpful in building regression models. For example, to build a model as: Y=β0+ β0X1+ β0X2+ β0X3+ β0X4+ β0X5

Observation (record) with any missing of X1- X5 will be excluded. If one X (e.g. X5) was removed from the model, how many more observations will be save dependents on the missing pattern.

A sample screen shot of input window:

Sample output and explanation:

Frequency of missing for each variable

Variable

# Non-missing

# Missing

SBP

795

37

DBP

795

37

SNP1

817

15

SNP2

814

18

BMI

795

37


Frequency of joint missing pattern for selected variables

SBP

DBP

SNP1

SNP2

BMI

Frequency

0

0

0

1

0

2

0

0

1

1

0

35

1

1

0

0

1

1

1

1

0

1

1

12

1

1

1

0

1

17

1

1

1

1

1

765

Created on 7/10/2011 with Empower(R) (www.empowerstats.com) utilizing R

The first table lists the number of missing for each individual variable. The table 2 shows the missing pattern and number of observations with such joint missing pattern. The 0 represents the variable is missing, 1 represents the variable is non-missing. E.g. 35 records had SBP, DBP and BMI missing and SNP1, SNP2 non-missing. 765 records had complete information (non-missing) for all the 5 variables.