The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » India » Comparing Indian states in multi-year NFHS (State-wise comparison India)
Comparing Indian states in multi-year NFHS [message #26017] Fri, 27 January 2023 16:25 Go to next message
Isha is currently offline  Isha
Messages: 18
Registered: February 2016
Member
Hello,

I am using the NFHS3, 4 and 5 couple datasets, and comparing states in terms of gender-related attitudes and practices. I do not intend to say something about a specific state but compare responses among the different states.

1. I am using national level weights (men's). I believe I do not need to use state-level weights since I am describing states in comparison to one another. I am using the svy command. I hope it is fine to use the national-level weights only?

2. There are some regions that are not in all three datasets, hence I will delete these from my analysis (eg. Andaman & Nicobar, Lakshadweep Islands, Dadra & Nagar Haveli, Daman & Diu). I hope that is okay.

3. However, some regions are overlapping (eg. Ladakh is separate in NFHS-5 but part of Jammu and Kashmir in NFHS3 and NFHS4). What to do in this case? Can I merge Ladakh and J& K in NFHS-5 for analysis and comparison with J & K with NFHS3 and NFHS4?
The same issue comes up with Telangana and Andhra Pradesh? Telangana would have been included in Andhra Pradesh in NFHS3 and NFHS4, but it is separate in NFHS5. I do not want to drop these big states: J& K and Andhra Pradesh.

4. What to say about regions like Chandigarh and Puducherry? Would NFHS3 have included these in Punjab or Tamil Nadu? Or do you think it is better to drop in NFHS5, or keep as missing data for NFHS 3?

Please advise!

Best,
Isha
PhD Candidate
Development Sociology
Cornell University
Re: Comparing Indian states in multi-year NFHS [message #26027 is a reply to message #26017] Mon, 30 January 2023 17:39 Go to previous messageGo to next message
fred.arnold@icf.com is currently offline  fred.arnold@icf.com
Messages: 75
Registered: May 2021
Senior Member
You should use the national level weights when comparing states. On 31 October 2019, the state of Jammu & Kashmir was reconstituted as two union territories, one of which was called Jammu and Kashmir, and the other of which was called Ladakh. For that reason, I would not recommend combining Ladakh and J & K in NFHS-5 to make comparisons with the earlier surveys. Telangana was made a separate state (carved out of Andhra Pradesh on 2 January 2014, one year before the start of the NFHS-4 fieldwork). Therefore, the state of Andhra Pradesh can be compared between NFHS-4 and NFHS-5, but neither NFHS-4 nor NFHS-5 can be compared with NFHS-3 for Andhra Pradesh. Telangana can be compared between NFHS-4 and NFHS-5, but it was not a state in NFHS-3. Chandigarh was a union territory in all three of the most recent NFHS surveys. It has never been a state. Chandigarh has an unusual relationship with Punjab and Haryana. Chandigarh is the capital of both Punjab and Haryana, but it is not part of either of those states. Jharkhand was separated from the state of Bihar and made a separate state on 15 November 2000, even before NFHS-3. Therefore, Jharkhand can be compared across all three surveys, and Bihar can be compared across all three surveys. Similarly, Chhattisgarh was separated from the state of Madhya Pradesh on 1 November 2000, which was before NFHS-3. Therefore, both Chhattisgarh and Madhya Pradesh can be compared across all three surveys.
The union territories of Puducherry, Andaman and Nicobar Islands, and Lakshadweep can be compared across all three surveys if you want to include them in your analysis. Although the name Pondicherry was changed to Puducherry in 2006, that union territory covered the same areas at the time of all three of the most recent NFHS surveys. The union territories of Dadra & Nagar Haveli and Daman & Diu were merged into a single union territory (Dadra & Nagar Haveli and Daman and Diu) on 26 January 2020. The combined union territory was included in the NFHS-5 national report, but the separate union territories were shown in the NFHS-4 national report. The two separate territories in NFHS-4 should not be compared with the combined union territory in NFHS-5.
Re: Comparing Indian states in multi-year NFHS [message #26562 is a reply to message #26027] Fri, 31 March 2023 15:26 Go to previous message
Isha is currently offline  Isha
Messages: 18
Registered: February 2016
Member
Dear Dr. Arnold,

Thank you so much for your helpful comments about comparing states in the various rounds of couple data of the NFHS. I have used your advice in my doctoral dissertation paper. The DHS is also an exciting dataset to work with and with many possibilities.

I really admire the research you do, and your work on fertility preferences. Thank you for inspiring us!

Best,
Isha
Previous Topic: Merging DHS and DLHS data - India
Next Topic: Wealth Index variables
Goto Forum:
  


Current Time: Fri Mar 29 04:07:27 Coordinated Universal Time 2024