The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » India » School attendance for 6-17 years
School attendance for 6-17 years [message #14004] Fri, 02 February 2018 05:19 Go to next message
neelanjana.p@gmail.com is currently offline  neelanjana.p@gmail.com
Messages: 1
Registered: January 2018
Location: New Delhi
Member
Hi all,
I am working on school attendance , just trying to replicate table 2.19 for Bihar
My estimate is not matching with reported 83.8 state level estimate using hv121, i am getting 85.62 . Am I using wrong variable for computation?
This is my code:

svy: tab hv121 if hv105>5 & hv105<18 & hv024==5 & hv103==1

Outcome
member |
attended |
school |
during |
current |
school |
year | percentages
----------+------------
no | 14.37
attended | 85.62
don't kn | .0186
|
Total | 100



Please help.

Re: School attendance for 6-17 years [message #14141 is a reply to message #14004] Fri, 23 February 2018 09:06 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3017
Registered: February 2013
Senior Member
Following is a response from Senior DHS Stata Specialist, Tom Pullum:


We apologize for the delay in replying to your February 2 question. DHS has a special procedure for this type of table on school attendance. To get the denominator, you need to know who was in the appropriate age range (e.g. 6-17) in the month of eligibility for school attendance (e.g. April of the calendar year in question). There are several potential ways to estimate whether someone should be in that denominator. The DHS approach is to merge the BR, IR, and MR files with the PR file, thereby obtaining a cmc for month of birth for as many cases as possible (using b3, v011, or mv011, respectively). For remaining cases for which you only have hv105 (age at last birthday in the PR file) to work with, you assume that the person is half way through the year of age. I looked at the NFHS-3 (IA52) data (the relevant table is 2.9) and found that it contains a variable sh18c that gives age in months in April of 2005. Unfortunately, that variable is not routinely put into the data files and it is not in the NFHS-4 (IA71) data (the relevant table is 2.19). I have written a Stata program to replicate the construction of sh18c in the NFHS-3. I do not have time to modify it for the NFHS-4 but you should be able to do that modification. You will of course have to change all the paths as well as the filenames. Also, in NFHS-4 there was separate treatment of the two phases of the survey, so you have to adapt to April of two successive school years. The ages 6 and 17 are included as scalars.
Re: School attendance for 6-17 years [message #14246 is a reply to message #14141] Sun, 11 March 2018 16:48 Go to previous messageGo to next message
mianrashid is currently offline  mianrashid
Messages: 13
Registered: February 2016
Location: Pau, France
Member
Can we get information about schooling outcomes of children like current enrolment of children in private school or govt school in DHS data set?

MianRashid
Re: School attendance for 6-17 years [message #14257 is a reply to message #14246] Wed, 14 March 2018 09:17 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3017
Registered: February 2013
Senior Member

Following is a response from Senior DHS Stata Specialist, Tom Pullum:

School attendance is included in the household survey, for all persons of school age in the specific country. See, for example, variables such as hv121 in the PR file. However, there is no information about the type of school, e.g., public vs. private.
Re: School attendance for 6-17 years [message #14258 is a reply to message #14257] Wed, 14 March 2018 09:26 Go to previous messageGo to next message
mianrashid is currently offline  mianrashid
Messages: 13
Registered: February 2016
Location: Pau, France
Member
Thanks for Reply, recently I participated in a conference, Where a presenter presented a paper and they used this information in her study and mentioned,
" The latest Pakistan Demographic and Health Survey (PDHS) 2012-13 provides data on both nutrition status and schooling outcomes of children. With the sample size of 13,400 households, PDHS provides nationally representative estimates.
Logistic regression will be used with dependent variable: type of school currently attending (government/private). The independent variable for the analysis is nutrition status of children: Underweight/ stunted/wasted). The list of control variables is as follows: wealth quintile, parental education, age of the child, gender, residence (urban/rural), and region (province).",
So, I was finding whether this information is available or not.


MianRashid
Re: School attendance for 6-17 years [message #14259 is a reply to message #14258] Wed, 14 March 2018 13:37 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3017
Registered: February 2013
Senior Member

Following is a response from Senior DHS Stata Specialist, Tom Pullum:


The report on the 2012-13 Pakistan survey (https://www.dhsprogram.com/pubs/pdf/FR290/FR290.pdf) includes the household questionnaire in Appendix F. On page 278 of the report you will see the questions (in English) about schooling. There is no question about whether the school attended is public or private. I also cannot find a variable in the data set which makes that distinction. Perhaps the reference was to the current DHS survey in Pakistan, which is just finishing data collection this month. I cannot easily check the questionnaire for that survey. It would be very unusual to include a question on the type of school. The data from the current survey will probably not be available until 2019.
Re: School attendance for 6-17 years [message #18157 is a reply to message #14141] Tue, 01 October 2019 07:28 Go to previous message
biswaj41_ssf is currently offline  biswaj41_ssf
Messages: 7
Registered: May 2019
Member
Modification according to NFHS-4 to replicate table no. 2.19 (India national report)

There are two phases in NFHS-4 and we need to define CMC separately for two phases. Just change to the following codes in the help file provided by Tom Pullum. No need to change other codes.

* Generating different CMC for two phases
gen CMC=.
replace CMC = 4+12*(2014-1900) if shphase == 1
replace CMC = 4+12*(2015-1900) if shphase == 2
tab shphase CMC
gen must_use_hv105=0
replace must_use_hv105=1 if in_BR==. & in_IR==. & in_MR==.
gen sh18c_test=CMC-cmc_of_birth

Enjoy!
Previous Topic: NFHS-4 Median Birth interval mismatch
Next Topic: Minimum Dietary Diversity Score
Goto Forum:
  


Current Time: Fri Mar 29 07:18:51 Coordinated Universal Time 2024