Reply With Quote
which s/w you are using? I guess you model coefficents also changed after doing the adjustment.
Hi There, I have a question regarding adjusted ODD RATIO interpretation
I am trying to interpret a table of results of logistic regression applied to
Environmental variables chosen as possible predictors of presence/absence of mosquito larvae in potential mosquito breeding sites.
In particular I am in doubt on how you interpret the adjusted odd ratios (OR) section especially in cases when
1) the Adjusted OR are a) very different from the crude OR b) when the adjusted OR indicate increased likelihood compared with the reference stratum (WINTER seems 2.13 times MORE likely than spring to favour the presence of larvae…..and this relationship seems to be opposite in the crude OR (WINTER seems 0.49 times LESS likely than spring to favour the presence of larvae). I have placed the data below for reference.
Many thanks for your help.
variable category code larvaepresent (no/%) larvae absent
Season Spring 1 19 (54.29) 16 (45.71)
Summer 2 2 (16.67) 10 (83.33)
Autumn 3 3 (12.00) 22 (88.00)
Winter 4 17 (70.83) 7 (29.17)
unadjusted OR
1 Spring p<0.0001
5.94 (1.13-31.15) Summer
8.7 (2.19-34.53) Autumn
0.49 (0.16-1.47) Winter
ADJUSTED OR
1 spring p=0.0005
7.23 (0.71 - 73.6) summer
32.14 (4.37 - 236.3) autumn
2.13 (0.41 - 11.15) winter
which s/w you are using? I guess you model coefficents also changed after doing the adjustment.
In the long run, we're all dead.
Posting Permissions
|
|