Question: Consider the following relationship between an exposure (E) and an Outcome (O) as depicted in the below 2x2 table:
|
|
Outcome
|
|
|
Exposure
|
Yes
|
No
|
Column Total
|
|
Present
|
30
|
34
|
64
|
|
Absent
|
24
|
23
|
47
|
|
Row Total
|
54
|
57
|
111
|
Calculate the Odds Ratio (OR) and interpret it. Now suppose a control variable Z is introduced and the partial 2x2 tables are provided below:
|
Low Z
|
|
|
Outcome
|
|
|
Exposure
|
Yes
|
No
|
Column Total
|
|
Present
|
13
|
8
|
21
|
|
Absent
|
20
|
10
|
30
|
|
Row Total
|
33
|
18
|
51
|
|
High Z
|
|
|
Outcome
|
|
|
Exposure
|
Yes
|
No
|
Column Total
|
|
Present
|
17
|
26
|
43
|
|
Absent
|
4
|
13
|
17
|
|
Row Total
|
21
|
39
|
60
|
Note that Z is non-causally related to the exposure (E) and is causally related to the outcome variable (O).
Is the variable Z a confounder or merely a control variable? Sketch the appropriate path diagram.