Survey data copying into other surveys

Question

With a recent update to Survey123, I had a problem with saved answers, wherein answers relevant to a previous answer was not being saved. I applied a workaround I was given and the problem was solved.

However, there is another problem I’m experiencing which is causing me a great deal of work. Answers from previous surveys seem to be bleeding into new surveys. I think it may be related to saved (Favorite) answers, but answers from one survey may be copying directly over into the next. Again, this has only been a problem since a relatively recent update.

My survey collects a variety of Wilderness management data from volunteers, including trail maintenance, campsite, and noxious weed data, etc.

As an example of this “bleeding:

* Surveyor begins the day by entering name, Wilderness, and trail (relevant to the wilderness selected), saves answers as favorites, and exits survey, saving it as a draft. I have recommended the old drafts be deleted before creating a new one.
* Surveyor begins hike and encounters deadfall on the trail. They paste the favorite answers and record this trail maintenance item as deadfall, and are presented with the questions about the deadfall (number, size, and whether they cleared it). They complete this, save the survey and move on.
* Surveyor next encounters a stream crossing. They paste the favorite answers and record this trail maintenance item as a stream crossing, and are presented with the questions about the crossing (type, condition, etc). They complete this, save the survey and move on.
* Surveyor next encounters noxious weeds. They paste the favorite answers and record this noxious weed infestation, and are presented with the questions about the infestation (species, phenology, extent, etc). They complete this, save the survey and move on.
* On uploading the surveys, the first survey is fine, but both the stream crossing and noxious weed surveys also include the information about the previous deadfall survey (number, size, and whether they cleared it).

I am having to go through each survey and edit out the unrelated data. For each survey that isn’t deadfall, I have to go in and delete the deadfall data. This “bleeding” is not limited to any one type of data. Noxious weed data may bleed over into a campsite survey.

Until the recent Survey123 update, the data editing required was minimal. Now it’s a major chore.

It has been my assumption that this was a problem with the saved answers, but perhaps not. It could be that answers from one survey are copying directly over into the next.

FOMELC01 · Answer

Unfortunately, I haven't been able to figure the problem out yet. I got on a Teams meeting with ESRI technical support  (Zach Sutherby), but he wasn't able to reproduce it and we got nowhere. His suggestion was to get the SQLite database from a survey from a user whose data was affected and send it to him so he could take a look at it. The problem was inconsistent and didn't affect everyone. By the time I had this suggestion it was field season and I simply didn't have time to get an affected SQLite file from someone with the problem (it never happened to my data). I got too busy to track down an affected SQLite file, so I just fixed the numerous errors as I went along (a bunch of work in itself), but never found a solution.

If you can get an SQLite file from someone with the issue , you might try submitting it to ESRI support with a description of your experience.

Zach said the workflow to extract the SQLite database from the users device is to have them open the files app on the device and navigate to the Survey123 folder.

* On IOS that will be under On my iPhone > Survey123
* On Android that will be <Device name> > Android > data > com.esri.survey123 > files

Once there select the ArcGIS folder > My Surveys > Databases. Within that Databases folder there will be a file with a GUID name. Pass along that file with the GUID name as that will be the SQLite database that will contain the submission.

I wish I could be more helpful and come up with a solution to the problem. I may yet be able to figure it out in the "slow" season. If so I'll post it.

Elizabeth_Burniston · Answer

@RalphBradt1 I am having the exact same problem, having my users follow the same the exact same process as you describe. At the beginning of the day, they enter a series of information, save as favorites, exits the survey, and save as draft. The person begins a new survey (not from draft... using collect), paste from favorites, enters the other information as appropriate and submits. This goes on throughout the day. The first survey is fine, but the subsequent ones have the bleed over affect from that first survey just as you describe. Was there anything you figured out? I'm up to 2,100 submissions so far and this is killing me.