Required geopoint being "skipped"?

[NickDierks]

We had an issue recently where a user was able to "skip" a required geopoint question in a survey. Apparently, simply opening the survey was enough to get an automatic rough point (usually with accuracies in the hundreds or thousands of meters), and no user input was required. Is this intended behaviour?

For what it's worth, there was a warning about bad accuracy, but it and the geopoint question were located in a closed group due to survey size. The user simply never bothered to open the group, for whatever reason. Moving forward, I know I can mitigate this issue in several different ways, but I was surprised that this happened at all. I figured a required geopoint would require actual user input before allowing a survey to be submitted.

[JamesTedrick · Accepted solution]

Hi Nick,

The default behavior of Survey123 is to capture the user's location.  Location, as provided by the phone, can be somewhat rough, especially if the GPS hasn't been recently used.   You can specify an accuracy threshold in the body::accuracyThreshold column (see Geopoints—Survey123 for ArcGIS | ArcGIS).

[SMauri]

Hi Nick,

My approach is to specify an accuracy threshold and/or to limiting of the extent in which points can be input into a survey.

So when device has a "poor" location inside, user is forced to request location by pressing button of GPS.

You can find more detail here: https://community.esri.com/groups/survey123/blog/2017/10/26/limiting-extent-in-which-input-point-can... 

Stefano

[Jean-YvesLandry1]

With the new 2.5 update you can  pulldata() from a picture EXIF,   what about asking them to take a picture then use that gps point as your geopoint value?

add yes to read only so they can't change the gps value.