How can I select observations containing specific words?

[Celine Tran]

Dear all,

I need to pick persons A & B whose position include exact "Independent" from the list as follow

Directors Position
A Non-executive Independent Director
B Independent Non-executive Director
C Non-executive Director
D Non-executive Chairman of Board
E Non-Independent Non-executive Director

How can I do that? Because if I use the command

gen InD=regexm(CurrentPosition,"Independent")

Person E is chosen too.

Thank you

[Nick Cox]

gen wanted = strpos(lower(CurrentPosition), "independent") > 0  & strpos(lower(CurrentPosition), "non-independent") == 0

That code corrects for variations in case but not otherwise for misspellings or abbreviations.

[Marcos Almeida]

This is an alternative including "Non-independent":

gen myvar = regexm(var1, "[Ii]ndependent")

Edited: sorry, this would unfortunately preserve the "Non-Independent" term.

[Marcos Almeida]

Heureka.

This command may cope with initials (lower or upper) as well:

gen myvar2 = regexm(var1, "[Ii]ndependent") & strpos(var1, "Non-Independent") ==0

[Celine Tran]

Thank you very much for your help !