Locating respondents for follow-up surveys

Support
1

I'm working on a project in a rural community where huts can be close to one another. In addition the names of the household head can sometimes just be a first name.

Both of these factors present a problem when ennumerators are returning to an area for a follow-up or interim survey. The problem being that how do we know ennumerators are interviewing the same household head? I read with interest this post on geofencing for ODK but again, the proximity of huts doesn't really make this feasible.

I would like to know what other solutions ODK users have come up with for this (presumably) common problem.

I'm not certain this belongs in support as it is more a survey strategy question than a design question but any ideas would be weclome.

We are using aggregate, collect 1.9.1 and XLSForm to create forms.

2

Hi Oliver! I agree that this is more of a survey strategy problem.

When you do the baseline survey, you need to collect as much identifying information as possible. So maybe it's a participant name, father's name, mother's name, children's names, birth date, etc. That information combined with the location is usually enough to identify the participant.

I've seen other campaigns where they use a national ID or leave an ID card for the person to keep. You can also take photos of participants, but then you are generating a lot of data that you might not use.

1 Like
3

Thanks Yaw. On longer surveys I was hoping to avoid respondent fatigue by repeating questions from previous surveys but this is helpful.

4

You don't have to repeat the questions. What I do is use pulldata and calculates to store the data from the baseline in the followup form, then show that hidden data in a read-only screen like so...

participant name: ${name}
participant age: ${age}
father name: ${father_name}

Then the enumerator just has to review that screen to confirm they have the correct participant.

5

We use a set of printed cards with uniqueID which we distribute to the respondents during the first survey. This physical card and number is used by the data collectors in the subsequent rounds to verify and pre-load information, and sync information between forms.

Note, the form filled during the first survey has an input for entering this uniqueID and all required identifying details of the household.

1 Like
6

Great idea. Thanks Vanubhav. Presumably QR codes could be used too.

7

QR or Barcodes would work great! But they're hard to replace, and in our particular situation, the farmers often misplace the cards, or leave it at home when we meet them in the fields. A number is simpler to remember. But there's no reason why you cant print both!

1 Like