-mrtab- on SSC might help. Type
. ssc install mrtab
to download the program. Then type
. mrtab concern1-concern3, poly response(1/6) countall sort
in order to see a table of counts for the 6 categories. It is
also possible to save count variable for the individual
categories:
. mrtab concern1-concern3, poly response(1/6) countall ///
generate(C)
will produce 6 variables C1-C6 which can then be further
processed by applying
. collapse ..., by(VID)
A solution to deal with ties would be to apply -mrtab-
repeatedly: first to concern1-concern3, then to
concern1-concern2, then to concern1 (or just use
-tabulate- for the last step).
Cheers,
ben
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of
> Farooq Naseer
> Sent: Friday, November 26, 2004 6:55 AM
> To: [email protected]
> Subject: st: Finding 3 most popular responses to a question
>
>
> Dear all,
>
> A randomly selected group of people within some survey
> villages were given the
> option to pick their top three every-day concerns: Of the six coded
> options available, a respondent could pick, say, (4,4,2) and another
> one in the same village could pick (1,6,5). (Duplicates like 4,4,..
> occur due to the fact that the responses were post-coded into six
> broad categories to make analysis simpler).
>
> Now i want to get a measure for the community's top three priorities
> by aggregating the information contained in the above individual
> responses.
>
> I have the above info in a stata dataset as variables:
> concern1-concern3. To avoid complications re. appropriate weighing
> scheme, for the time being, I just want to take an unweighted count of
> the values in these 3 variables for each village -- the variable VID
> -- in my sample. In case there is a tie, for instance a village 'j'
> has 20 responses each for values 4 and 6, I want to pick the value
> which gets mentioned more in a higher-priority variable acc. to the
> following priority ranking: concern1 > concern2 > concern3.
>
> I know it should be straight-forward to handle in Stata. But after
> having spent some on it, and seeking help from Stata's Programming
> manual, the program I produced is not working correctly and is quite
> inelegant in its use of all the nested loops. I attach the program, as
> is, below. Any help will be (greatly)^greatly appreciated!
>
> Thanks!
> Farooq
>
> **************
> set trace on
> set more off
>
> bysort VID: gen N = _N
>
> local j = 1
> local i = 0
>
> while `j' < _N {
>
> local s1_x1=0
> local s1_x2=0
> *... likewise i define a total of 18 such local macros -- s1_x1 -
> s3_x6 -- to tempo-*rarily store the count for all 6 options in each of
> the three concern vars.
> *...
> forvalues k=1(1)3 {
> local i = `j'
> while `i' < `j'+N[`i'] {
> if concern`k' == 1 in `i' {
> local s`k'_x1 = `s`k'_x1' + 1
> }
> else if concern`k' == 2 in `i' {
> local s`k'_x2 = `s`k'_x2' + 1
> }
> * removed similar else if statements for concern`k'==3, 4, 5
> & 6 to save space
> *...
> else {
> * do nothing
> display "Missing Value, Do Nothing"
> }
> local i = `i' + 1
> }
> }
>
> * Now i have the count for all occurences of concern by value in each
> of the top 3 concerns. Need to take an unweighted sum to get top 3 and
> resolve ties.
>
> local s_x1 = `s1_x1'+`s2_x1'+`s3_x1'
> local s_x2 = `s1_x2'+`s2_x2'+`s3_x2'
> * and so on... until s_x6...
>
> local j = `j' + N[`i']
> }
> *- the outer j while loop
> * haven't finished the entire program yet but the results so far don't
> seem correct...
>
> set trace off
> set more on
> *
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
