| Missing variables in data set [message #10559] |
Mon, 08 August 2016 09:29  |
Peter M
Messages: 1
Registered: August 2016
Location: Nairobi
|
Member |
|
|
Greetings everyone.
I am using the 2014 KDHS to analyse tobacco use in the country but have realized the data set on individual and household member recode have no variables on smoking i.e. V463 A-E. The only one with the variable is the couples recode. Am lost on what to do as these variables are captured in the DHS. Also, the KDHS interviewed more than 36000 households but the I am anly able to see less than 6000 responses from the recode files downloaded.
Anyone who can advise what am missing?
Many thanks in advance.
Peter
POM
|
|
|
|
| Re: Missing variables in data set [message #10562 is a reply to message #10559] |
Mon, 08 August 2016 12:34  |
Bridgette-DHS
Messages: 3196
Registered: February 2013
|
Senior Member |
|
|
Following is a response from DHS Senior Research Associate, Shireen Assaf:
I am using the file KEIR70FL.dta and I have found observations in the variables on smoking from v463a-e. Please be sure you have the correct file. Also please be aware that for the 2014 KDHS there was a long and short questionnaire and some questions were only asked in the long questionnaire. Therefore, you could see fewer observations for some variables. Please see Appendix E for a description of what was included in the long and short questionnaires and also read section 1.7 of the 2014 DHS final report http://www.dhsprogram.com/pubs/pdf/FR308/FR308.pdf.
To know which variables in the dataset were in the long or short questionnaire, you must use the variable hv027 from the PR file. Please merge the PR and IR files. If you select for hv027==1, this is the long questionnaire since they were selected for the men's interview.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
