st: Encode string variables without following the default alphanumeric ordering

[Adama Konseiga <akonseiga@gmail.com>]

Dear Statlisters

I am dealing with a questionnaire where one question leads to rating
an activity on a scale from 0 to 10.

There are about 25 variables from that question which are currently in
a string format, including missing and text such as "0 - Activity not
interesting at all".

I engaged in automatizing the process of encoding the 25 variables
into numeric. But neither -encode- nor super encode are doing the job
properly.
They are based on alphanumeric order of appearance of the string
values, which leads to numeric values having completely different
meanings accross the variables.

Does anyone has some suggestions how to improve my following codes.

-------

Stata version: Stata 10.1

u mydataset, clear

/*
tab activitym1, miss

br activitym1
egen tag=tag(activitym1)

list activitym1 if tag
****Confirm that the order of apperance of string values are different
in the 25 variables
*/

foreach var of varlist activitym* {
sencode `var', replace label(scores, replace)
 }
recode activitym1 (1=8) (2=5) (3=0) (4=6) (5=9) (6=2) (7=10) (8=7)
(9=4) (10=3) (11=1)
.....
-----

recode activitym4 (1=0) (2=2) (3=6) (4=7) (5=8) (6=5) (7=9) (8=10)
(9=1) (10=3) (11=4)


-- 

Adama Konseiga
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/