Business Context
In a clinical drug trial, some patients dropped out, resulting in missing values for the 'SideEffects' variable. The researchers must include these missing values in the analysis to determine if the occurrence of missing data is associated with a specific 'TreatmentGroup' (Placebo vs. Active).
Data Preparation
Generating trial data with missing values in SideEffects.
Copied!
| 1 | |
| 2 | DATA casuser.clinical; |
| 3 | call streaminit(777); |
| 4 | DO i=1 to 200; |
| 5 | IF mod(i,2)=0 THEN Group='Placebo'; |
| 6 | ELSE Group='Active'; |
| 7 | r = rand('UNIFORM'); |
| 8 | IF r < 0.15 THEN SideEffects=' '; |
| 9 | ELSE IF r < 0.6 THEN SideEffects='None'; |
| 10 | ELSE SideEffects='Nausea'; |
| 11 | OUTPUT; |
| 12 | END; |
| 13 | |
| 14 | RUN; |
| 15 |
Étapes de réalisation
1
Execute crossTab with includeMissing=true and association=true.
Copied!
| 1 | |
| 2 | PROC CAS; |
| 3 | SIMPLE.crossTab / TABLE={name='clinical', caslib='casuser'} row='SideEffects' col='Group' includeMissing=true association=true; |
| 4 | |
| 5 | RUN; |
| 6 |
Expected Result
The crosstab results must contain a row for missing 'SideEffects' (blank or dot). The 'Measures' table should be generated, providing statistics (like Cramer's V) to quantify the association between the treatment group and the side effects (including the missing ones).