The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Kenya » KDHS 2014: Table 2.14 School attendance ratios
Re: KDHS 2014: Table 2.14 School attendance ratios [message #28143 is a reply to message #28075] Fri, 17 November 2023 11:27 Go to previous messageGo to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3202
Registered: February 2013
Senior Member
Following is a response from Senior DHS staff member, Tom Pullum:

I have added a Stata program that is an improvement on the one you were using and on the GitHub program, although it just calculates the net attendance ratios in table 2.14 and just by the child's sex and place of residence. The other parts of table 2.14 should be easy to add. I had written the Stata program now on GitHub several years ago, basically as a translation of a CSPro program originally written by Trevor Croft long ago. The attached version has some simplifications and more comments.

In the original CSPro program for this table for the Kenya 2014 survey, the year and month for eligibility are 2014 and 2 (February), respectively. You were using month 1 (January). It is quite possible that month 1 would have been more consistent with the Kenyan school system, but the program used month 2. In any application of this program, it is essential to adjust those numbers as well as the age ranges for primary and secondary school.

A peculiarity of the DHS procedure (which probably originates with a UN agency) is that a child who is primary age but attending school is counted as not in school. Similarly, a child who is secondary age but attending post-secondary is counted as not in school. I don't think this is fair to kids who manage to skip grades, but that's the procedure.

As stated in the title, the procedure is limited to children who are de facto residents of the household, i.e. were in the household the previous night. This makes some difference.

Unfortunately, the procedure includes a random component that makes it impossible to match the table. The child's cmc of birth is given by b3 if the child is in the BR file, as well as the PR file. For children who are in the PR file but not in the BR tile (and there are many of them), the month of birth is imputed with a uniform random distribution that is consistent with the stated age in years (hv105). That means different results will be given by different random number generators and different seeds. My program just uses "uniform()", and the table in the report uses the CSPro generator. There is no way around this issue. It would be possible to produce a version in which all children were assigned to month 6, or to month 7, which would avoid the random component but would not be as accurate. The bottom line is that you can never match the table on school attendance ratios exactly.



Report message to a moderator

[Message index]
 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: KDHS 2022: Table 2.12 School attendance ratios
Next Topic: SVYSET
Goto Forum:
  

[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Mon Dec 2 15:33:17 Coordinated Universal Time 2024
.:: Contact :: Home :: Privacy Policy ::.

Powered by: FUDforum 3.0.9.
Copyright ©2001-2018 FUDforum Bulletin Board Software

The information provided on this Web site is not official U.S. Government information and
does not represent the views or positions of the U.S. Agency for International Development or the U.S. Government...