The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Biomarkers » Identifying Interviewers who collected biomarker for women/men/children (How to construct variable to identify which investigator collected biomarker response for respondent)
Identifying Interviewers who collected biomarker for women/men/children [message #19680] Thu, 30 July 2020 11:59 Go to next message
kajori is currently offline  kajori
Messages: 5
Registered: May 2017
Location: Mumbai, India
Member
Hello,

I am using the Person file in NFHS 4, 2015-16 to identify teams that have collected biomarker samples from respondents.

Using state codes and supervisor codes in the following command:

egen interview_team=group(hv024 hv030)

in Person file I am trying to identify the groups that have worked in states in NFHS 4. However, according to report 789 teams have worked. Using this command I am getting 875 field teams. Can you please let me know how can I identify the 789 field teams using NFHS data?

Thanks in advance.

[Updated on: Thu, 30 July 2020 12:00]

Report message to a moderator

Re: Identifying Interviewers who collected biomarker for women/men/children [message #19702 is a reply to message #19680] Tue, 04 August 2020 13:41 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1850
Registered: February 2013
Senior Member
Following is a response from DHS staff involved in the fieldwork and data processing of the NFHS 4 survey:

There are indeed 875 unique state/team combinations in the NFHS4 data files. However, it is probably impossible to know the correct number, because some teams had more than one ID code name. During fieldwork, the field agencies were moving people around. There are definitely instances of the same interviewers having different IDs based on what teams they were reassigned to.

For example, the state of Telangana was part of Andhra Pradesh at the time of data collection. It is likely that the 22 teams from AP and TG are the same teams. Most likely there are no teams with different IDs within each state, but it's hard to tell which teams/interviewers were shifted between states and assigned new numbers.

The team ID number is based on the supervisor, so even if the supervisor had been reassigned from one team to a new one, it doesn't mean all the old team members were also reassigned to the same new team.

It's best to regard the number in the report, 789, as an estimate.

[Updated on: Tue, 04 August 2020 14:09]

Report message to a moderator

Re: Identifying Interviewers who collected biomarker for women/men/children [message #19703 is a reply to message #19702] Tue, 04 August 2020 14:33 Go to previous messageGo to next message
kajori is currently offline  kajori
Messages: 5
Registered: May 2017
Location: Mumbai, India
Member
Thank you for the response.
However, we are trying to assess team level effects in some data estimates from NFHS 4, 2015-16. To do this we need to identify teams in the data file (person file/ household file). Can you please suggest on how a team level identifier can be generated using the information provided in the person file?
Re: Identifying Interviewers who collected biomarker for women/men/children [message #19709 is a reply to message #19703] Tue, 04 August 2020 18:05 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1850
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

I recommend that you use the ID codes as they exist in the data. You really have no alternative.

For this survey you have evidence that there was some inconsistent use of interviewer ID codes, but I believe that in most surveys there are similar issues even if they are completely hidden. The ID codes for interviewers almost always follow a pattern, which I described in a 2018 Methodological Report (https://www.dhsprogram.com/pubs/pdf/MR24/MR24.pdf). The final digit in the ID code usually describes the position of the person on the team. The final digit is "0" for the supervisor. Certain digits are reserved for the measurers. A digit may be reserved for a person (male) who conducts interviews of men.

Suppose that during fieldwork an interviewer, for whatever reason, must leave the team. That person will be replaced and the new person will take over the same ID code. Other substitutions can happen and lead to the same ID code being used by more than one interviewer.

Although these things happen, they are relatively rare. With the NFHS4, the two numbers (875 and 789) are within about 10% of each other. I would take the data at face value, but would add a footnote or comment that the there is evidence of some inconsistent use of ID codes.



Re: Identifying Interviewers who collected biomarker for women/men/children [message #19722 is a reply to message #19709] Wed, 05 August 2020 10:36 Go to previous message
kajori is currently offline  kajori
Messages: 5
Registered: May 2017
Location: Mumbai, India
Member
Thank you! I have read the methodical report also. So, I will proceed with the data at its face value and make a note of the difference. Thank you so much for the help.
Previous Topic: STAT Compiler
Goto Forum:
  


Current Time: Mon Aug 10 01:34:28 Eastern Daylight Time 2020