I know how to calculate the odd ratio by using the 2x2 table. But how do you calculate the adjusted odd ratio in 2x3 table, either manually or with software?
SPSS or any other stats software will do (I only mention SPSS because the first poster mentioned it - I prefer SAS).
When you say a 2x3 do you mean that your outcome is binary but you have a predictor or exposure that has three categories? If so, you'll need to make a decision about whether you want it treated as an ordinal variable or whether you want to make a series of pair-wise comparisons (i.e. assign one value to be reference and compare the other two values to it). Different programs do this differently but it is something that you'll have to decide. Depending on how it is coded and what you choose it might be as simple as entering it into the model as is. Or you might have to specify the reference or format within the model (proc logistic within SAS has some options)
When you say "adjusted odds ratio" do you mean that there are also other covariates that you want to adjust for in the model?
You can use logistic regression for calculating the odd ration by using SPSS
SPSS or any other stats software will do (I only mention SPSS because the first poster mentioned it - I prefer SAS).
When you say a 2x3 do you mean that your outcome is binary but you have a predictor or exposure that has three categories? If so, you'll need to make a decision about whether you want it treated as an ordinal variable or whether you want to make a series of pair-wise comparisons (i.e. assign one value to be reference and compare the other two values to it). Different programs do this differently but it is something that you'll have to decide. Depending on how it is coded and what you choose it might be as simple as entering it into the model as is. Or you might have to specify the reference or format within the model (proc logistic within SAS has some options)
When you say "adjusted odds ratio" do you mean that there are also other covariates that you want to adjust for in the model?
concerning the crude OR and adjusted OR. When you estimate only on factor such as sex on infection you can get COR but when use more than one you can get AOR:
Bellow an example submitted by SAS.
data f;
input no sex age TH BAB ANA SE;
if age lt 1.1 then age=1 ;
if age gt 1 then age =2;
if se=3 then se=2;
if se=4 then se=1;
cards;
1 1 2 0 1 0 4
2 1 1 0 1 0 4
3 1 1 0 1 0 4
4 2 1.6 0 1 0 4
5 2 2 0 1 0 4
6 2 1 0 0 0 4
7 2 1 0 0 0 4
8 2 1 0 0 0 4
9 2 1 0 0 0 3
10 2 1 0 0 0 3
11 2 1 0 0 0 3
12 2 1 0 0 0 3
13 2 1 0 0 0 3
14 2 1.6 1 1 1 3
15 2 2 0 0 0 3
16 2 2 0 0 0 2
17 2 1 0 0 0 2
18 2 1 0 0 0 2
19 2 1.8 0 0 0 2
20 2 1.8 0 0 0 2
21 2 1.8 0 0 0 2
22 2 1.8 0 0 0 2
23 2 1 0 0 0 2
24 2 1.4 0 0 0 1
25 2 1 0 0 0 1
26 1 1.4 0 0 0 1
27 1 1 0 1 0 1
28 1 1.4 0 0 0 1
29 2 1 1 0 0 1
30 1 1.4 0 0 0 1
31 1 1.6 0 0 0 2
32 1 1.6 0 1 0 2
33 1 1.6 1 0 0 2
34 2 1.6 1 0 0 2
35 2 1.6 0 0 0 2
36 2 1.6 0 0 0 3
37 2 1.6 0 0 0 3
38 2 1.6 0 0 0 3
39 2 1.6 0 0 0 3
40 2 1 0 0 0 3
41 1 1 1 1 1 2
42 1 1 1 1 1 1
43 2 2 1 1 0 3
44 2 2 1 1 0 1
45 2 1 0 0 0 1
46 2 1 0 0 0 1
47 2 1 0 0 0 1
48 2 1 0 0 0 2
49 2 1 0 0 0 2
50 2 1 0 0 0 2
51 2 1 0 0 0 2
52 2 1 0 0 0 4
53 2 1 0 0 0 4
54 1 1 0 0 0 4
55 1 1 0 0 0 4
56 2 1 0 0 0 4
57 2 1 0 0 0 4
58 2 1 0 0 0 4
59 1 1 0 0 0 4
60 1 2 0 0 0 4
;option ps=2000 ls=70 nodate nonumber;
PROC LOGISTIC DATA=f DESCENDING;
MODEL th = sex age se;
RUN;
PROC LOGISTIC DATA=f DESCENDING;
MODEL th = sex ;
RUN;
Awesome read! Thanks for all your tips around adjusted odds ratio.
- Badges
- Science topic