The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Dataset use in SPSS » Compute variables in Male datasets, KgDHS-2012
Compute variables in Male datasets, KgDHS-2012 Sun, 25 January 2015 14:19
 Valko Messages: 12Registered: October 2014 Member
Hello. I am working with the male datasets of the Kyrgyzstan DHS-2012. For my sample consisting of 460 married men I would need the total number of living sons and living daughters. To calculate these sums, in SPSS i use Compute Variables for a new variable for "a total number of living sons" with the following equation: MV202 (sons at home) + MV204 (sons elsewhere).

However, when i took the 2d group (married men who have one son), and manually summarise the number of sons at home with the number of sons living elsewhere, the amount does not correspond to the results in SPSS frequency table. Sons at home of 169 + sons elsewhere of 41 should give 210, however in the frequency table for living sons there is 167. Pls see below.

In this regards, I would appreciate for your advice in two questions:
1) whether my above steps were right to get a total number living sons as a sum of MV202 (sons at home) and MV204 (sons elsewhere)?
2) why there is a different result for living sons?

best regards,

Re: Compute variables in Male datasets, KgDHS-2012 [message #3716 is a reply to message #3675] Fri, 30 January 2015 09:59
 Bridgette-DHS Messages: 3094Registered: February 2013 Senior Member
Someone is reviewing your post, and we'll get back to you soonest.

Thanks
Re: Compute variables in Male datasets, KgDHS-2012 [message #3717 is a reply to message #3675] Fri, 30 January 2015 11:33
 Bert-DHS Messages: 2Registered: January 2015 Member
Hello Valko,

I do not have your sub-sample, but like you, I created a variable 'SONSLIV' for the total Kyrgyztan male data file using COMPUTE and produced the attached frequencies for MV202, MV204 and SONSLIV in Excel. I found 590 men with 1 child alive.
The crosstab of MV202 and MV204 (also attached) gives 526 men with 1 son at home (and none elsewhere) and 64 with 1 elsewhere (but none at home). 526 + 64 gives 590, which matches the number in 'SONSLIV'.

Thanks,

Albert Themme
• Attachment: OUTPUT1.xls
Re: Compute variables in Male datasets, KgDHS-2012 [message #3956 is a reply to message #3717] Tue, 10 March 2015 10:16
 Valko Messages: 12Registered: October 2014 Member
Dear Albert,

hello. thank you very much for your responce. Sorry, i could not give feedback earlier. as per your advice, i also used a command "Compute variables" on the whole sample of men. I agree that what you wrote is workable for men with sons living elsewhere and sons living at home having one son. I also received total living sons 590. But what about men who have two sons, three sons?
For example, 339 men having 2 sons at home and zero elsewhere plus 17 men having two sons elsewhere and none at home (339+17) would give us 356 men. However, in the Sonslive it is 404. maybe i did something wrong?

thank you again.
Re: Compute variables in Male datasets, KgDHS-2012 [message #3983 is a reply to message #3956] Sun, 15 March 2015 17:35
 Valko Messages: 12Registered: October 2014 Member
Dear Albert,

I clarified uncenrtainties with the sum of MV202 and MV204. Two Sons at home (339) plus one son elsewhere and one son at home (48) plus two sons elsewhere and no sons at home (17) give 404 which matches the number on Sonslive.

best regards.
 Previous Topic: Sample size and systematic sampling Next Topic: Individual Health Insurance Data - Rwanda
Goto Forum:

Current Time: Sun Jun 23 07:06:20 Coordinated Universal Time 2024