Hi All,
I have a 2^2 factorial design where both my factors are categorical and my response is a proportion. Is it possible to analyse such a response using factorial methods?
For each of the 4 factor combinations how does one statistically determine a "winner"?
Some more details, I have 2 factors A and B each taking levels a1, a2, b1, b2 and for each combination of factors a number of people receive the stimulus and of those a subset respond to it (my response is this proportion). In a table this would be
Run A B #getting stimulus #respond stimulus p(respond)
1 a1 b1 s1 r1 r1/s1
2 a2 b1 s2 r2 r2/s2
3 a1 b2 s3 r3 r3/s3
4 a2 b2 s4 r4 r4/s4
I am looking to maximise p(respond)
Any suggestions or references on this would be most appreciated.
Best regards,
Mary
|