| Dataset updates and version control [message #27278] |
Wed, 12 July 2023 16:44  |
MonicaG_gutt
Messages: 1
Registered: July 2023
|
Member |
|
|
|
I'm interested in learning more about changes/updates made to existing datasets and how those changes are documented (if at all). For example, I previously downloaded the Afghanistan 2015 individual recode dataset (filename = "AFIR70FL.DTA"). I see online it is now available to download as filename == "AFIR71FL.DTA". Is there a document that notes any changes made to the dataset between version 70 and 71? This question would apply to other previously downloaded datasets also. Thank you.
|
|
|
|
| Re: Dataset updates and version control [message #27340 is a reply to message #27278] |
Thu, 20 July 2023 10:15  |
Janet-DHS
Messages: 893
Registered: April 2022
|
Senior Member |
|
|
We do not document the changes between release version 0 (preliminary) and release version 1, as there are a myriad of small changes, but most of them are around the documentation (written notes and labeling of variables), and not the data content. Typically nothing that is used to calculate major indicators such as mortality, fertility, family planning, and the like, nor any of the weights are changed.
For all data releases after version 1, a data alert is usually appended to the documentation (ccIRvv.docx), explaining the changes made. That alert is also emailed to all users who downloaded the previous version.
We do recommend downloading the release version 1 when it is released, just to be sure that you are working with the most recent version. For most later versions, the changes are typically just to one or two variables and it is up to the user s to decide whether it would affect their analysis.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
