|
|
Re: Nutritional status of women [message #3917 is a reply to message #3550] |
Fri, 06 March 2015 05:18 |
Liz-DHS
Messages: 1516 Registered: February 2013
|
Senior Member |
|
|
Dear User,
Here is a response from one of our experts, Dr. Tom Pullum:
Quote:The main reason you have not matched table 11.10 on page 183 of the Nepal 2011 report is that you did not apply all the restrictions given in the footnote to the table. You need to remove the cases with v438 coded in the 9000's; the women who are pregnant; and the women with a birth in the preceding two months. You missed the third restriction, which requires looking at v008-b3_01. Here, v008 is the cmc of interview and b3_01 is the cmc of the most recent birth. You exclude the case if that difference is 0 or 1. I will put my modification of your code below and then the tabulation that corresponds to the bottom row of table 11.10.
gen bmi=v445/100
gen bmic=1 if bmi<18.5
replace bmic=2 if bmi>=18.5 & bmi<25
replace bmic=3 if bmi>=25 & bmi<30
replace bmic=4 if bmi>=30 & bmi<50
label define bmic 1"Underweight" 2"Normal" 3"Overweight" 4"Obese"
label values bmic bmic
gen ht_flag=0
replace ht_flag=1 if v438>9000
gen preg_flag=0
replace preg_flag=1 if v213==1
gen months_since_last_birth=v008-b3_01
gen recent_birth_flag=0
replace recent_birth_flag=1 if months_since_last_birth<=1
tab bmic if ht_flag==0 & preg_flag==0 & recent_birth_flag==0 [iweight=v005/1000000]
bmic | Freq. Percent Cum.
------------+-----------------------------------
Underweight | 1,055.0565 18.19 18.19
Normal | 3,963.9914 68.35 86.54
Overweight | 650.396456 11.21 97.76
Obese | 130.142492 2.24 100.00
------------+-----------------------------------
Total | 5,799.5868 100.00
|
|
|
|
Re: Nutritional status of women [message #12141 is a reply to message #12133] |
Fri, 31 March 2017 17:35 |
Liz-DHS
Messages: 1516 Registered: February 2013
|
Senior Member |
|
|
A response from Dr. Tom Pullum:
Quote:
Your Stata code should work. The discrepancy in the number of cases is small and may be due to the difference between de facto and de jure residence (see hv102 and hv103 in the household file, v134 in the IR file). Or it could be the difference between the women who are in the PR file and those who are in the IR file. These numbers may differ slightly. Please look into those possibilities. If you still can't get a match, please tell us.
[Updated on: Fri, 31 March 2017 17:35] Report message to a moderator
|
|
|
|
|
|
|
|
|
|
|
|
Re: Nutritional status of women [message #24714 is a reply to message #24701] |
Mon, 27 June 2022 15:23 |
Janet-DHS
Messages: 899 Registered: April 2022
|
Senior Member |
|
|
The main difference between the two approaches is that the second one explicitly drops cases that are outside a plausible range from 1200 to 6000. The first one does not drop those cases. Also, I the second approach assigns an NA code (".") to the excluded values, whereas the first approach uses "keep if", which may have worked ok here but is risky because it drops cases and changes the file.
When comparing two alternative recodes, or even when validating one recode, I recommend a cross tabulation that includes NA cases. Here, for example, you can do "tab BMI bmic,m" to see the difference. The second recode (bmic) is correct because it drops cases outside the plausible range.
|
|
|