Issues with label

[Ashish Bandhu]

Dear all,

I have a dataset having the variable clear_fp_info, about the information provided for the family planning method. I tried for encoding it, works well, after encoding, I found an unnecessary category of blank ones (due to skip in questions). Any help would be highly appreciated about why an extra category ( in my example 8) is coming up for blank cells.

Sample date is

* Example generated by -dataex-. To install: ssc install dataex
clear
input str16 clear_fp_info
" "               
"clear"           
"somewhat_clear"  
"somewhat_clear"  
"somewhat_clear"  
"clear"           
"clear"           
"clear"           
"clear"           
"clear"           
"clear"           
"clear"           
"clear"           
" "               
"not_at_all_clear"
"not_clear"       
" "               
"not_clear"       
" "               
"clear"           
"clear"           
"clear"           
"somewhat_clear"  
"somewhat_clear"  
"somewhat_clear"  
" "               
"somewhat_clear"  
"clear"           
" "               
" "               
"clear"           
"somewhat_clear"  
"somewhat_clear"  
"-99"             
" "               
"somewhat_clear"  
"somewhat_clear"  
"somewhat_clear"  
"clear"           
" "               
" "               
" "               
" "               
" "               
" "               
"somewhat_clear"  
" "               
" "               
" "               
"not_clear"       
" "               
" "               
" "               
" "               
" "               
"somewhat_clear"  
" "               
" "               
" "               
" "               
" "               
" "               
" "               
" "               
" "               
" "               
"not_at_all_clear"
" "               
" "               
" "               
" "               
" "               
"clear"           
" "               
"clear"           
" "               
" "               
" "               
"clear"           
" "               
" "               
" "               
" "               
" "               
" "               
" "               
" "               
"somewhat_clear"  
"clear"           
"clear"           
" "               
"clear"           
"clear"           
" "               
"clear"           
"clear"           
"clear"           
"clear"           
"clear"           
"clear"           
end

The syntax that I have used,

label define choices_clear_fp_info 1 "very_clear" 2 "clear" 3 "somewhat_clear" 4 "not_clear" 5 "not_at_all_clear" 6 "-88" 7 "-99"



encode clear_fp_info, gen(clear_fp_infoV2) lab(choices_clear_fp_info)
order clear_fp_infoV2, after(clear_fp_info)
drop clear_fp_info
rename clear_fp_infoV2 clear_fp_info


label list choices_clear_fp_info
choices_clear_fp_info:
1 very_clear
2 clear
3 somewhat_clear
4 not_clear
5 not_at_all_clear
6 -88
7 -99
8
Thanks

Ashish

[Nick Cox]

replace spaces with empty strings; then encode knows to ignore any plain missing values.

Other way round, -88 and -99 may be informative codes meaning something like "not applicable" or "refused to answer". For those, use extended missing values such as .a or .b and supply value labels.

In Britain at least 99 has a positive meaning for many people: https://en.wikipedia.org/wiki/99_Flake

[Ashish Bandhu]

replace spaces with empty strings; then encode knows to ignore any plain missing values.

Other way round, -88 and -99 may be informative codes meaning something like "not applicable" or "refused to answer". For those, use extended missing values such as .a or .b and supply value labels.

In Britain at least 99 has a positive meaning for many people: https://en.wikipedia.org/wiki/99_Flake

Thank you