Greetings,
On one of our project, we were skimming some of the forms submitted for quality control and doing some minor edits (correcting typos, translating open end questions), we noticed 46 forms have been duplicated and resubmitted on 31 July, which is not a date our field enumerators were working on (all data collection activities completed on 28 July). These 46 duplicated forms do not have an audit csv file.
I am attaching the below excel sheet with the original forms and duplicated forms (after deleting all respondent data and keeping only auto generated columns).
Could you please advise if possible what went wrong, and how to mitigate such instances in the future ?
Duplicates - 6 Aug 24.xlsx (22.6 KB)
Hi @archibald,
Please share your Excel sheet.
1 Like
Apologies, thought it had been uploaded. Shared now, thank you.
Hi @archibald,
My guess would be that this is not a mistake from your side and rather a server side bug. I checked your data and I am not sure if it’s my Excel or not but the start and end metadata do not show the times, only the dates, played around them to try to figure out which ones are “twins” and can confirm that there are, but I would like to make sure that they are duplicate submission, aka the datetime stamps in their metadata are exactly the same. That might help in understanding what went wrong as perfectly matching stamps from the same device is the sign of a problem that would not be from you, while stamps that are not similar (but you are sure that they are duplicates) is a sign that maybe it was a mismanipulation from your side.
Could you clarify that? Maybe a simple screen shot of two duplicate rows with the full date and time could help a lot here.
@archibald, I quickly scanned your shared document but did not find any duplicates. I checked the _id as well as the _uuid. Did I miss it somewhere?
Thanks @Kal_Lam
The duplications are the forms that were submitted on 31 July 2024. i deleted all the data related to the respondents answers to adhere to data privacy policy, but the fact is that all the data submitted on 31 July are actually duplicated forms, as we did no data collection that day, and all of the forms had the exact same data of the submissions prior to that day.
What is weirder is that they carried new _id and _uuid entries, almost seems like they were new submissions on their own, where in fact they contained the exact same data of prior submissions, but were missing audit.csv files.
Thanks @kevin_razafindrabe
The time stamps are actually present in this excel sheet, you can see it in the formula bar of each cell or you can just change the format to time and it will be visible in all cells.
