Dear all,
I am confronted with an analysis of data on the awareness of
measles and its causes. The responses are 'yes or no' to all
questions .We are interested in only 12 questions since the others
are not important.
For each question , you get the following summary table:
Yes No Total
a b a+b
where a and b are cell frequencies.
Initially I had computed a chi-square test to see whether there was
statistically significant difference between the two proportions
before commenting. For 12 questions I did 12 chi-square tests.
I further did a sub-group analysis to compare the proportions of
responses among those with 'no primary education, high school
and university education'-- which also required 12 tests.
I don't seem to be happy with this analysis at all.
Is there any method which could be used to compare the responses on
an overall measure of awareness? Can we derive a variable to measure
overall awareness? The study was one-off for each question/response.
Please assist.
Yours sincerely,
June E.Simakani
Department of Statistics
University of Transkei
South Africa.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|