How to track your field interviewers: progress, quality and suspicious responses
In fieldwork the problem is rarely the questionnaire. It is knowing what the team is doing: how many surveys each interviewer has done, whether quotas by area, age or gender are on track, and whether every response is real. When that is tracked with a spreadsheet each person fills in at the end of the day, problems show up late, when there is no budget left to go back out.
This guide is for research agencies, research teams, local governments, NGOs and companies running in-person surveys with several people.
1. One link per interviewer
First, every response should arrive with the code of whoever collected it, without the interviewer typing it (or typing it wrong). The simplest way is a hidden field in the link:
https://surviqapp.com/en/s/your-survey?interviewer=E07&area=north
Each interviewer opens their own link on a phone or tablet. Every response they collect is tagged with their code and area, and the respondent never sees it.
Step by step to build the links:
How to create hidden fieldsUse codes (E01, E02) instead of names. If you later share results with a client, you don't expose your team.
2. See each person's progress during the day
With the hidden field, you filter results by interviewer and see how many responses each one has, when they were collected and how their group answers. Compare by area and you see which areas are behind on quota.
Checking mid-morning and at the end of the day changes everything. If an interviewer has zero at 11 a.m. on Monday, you call them that day, not on Friday.
3. Quotas: track what you will need at the end
If the study needs, say, 50% women and 30% over 50 in each area, those questions should come first in the survey. Then you can check progress by quota every day and move interviewers before an area fills up with a single profile.
To set your quotas:
Calculate responses needed per group4. Spot suspicious responses
Almost every field team has a case of surveys filled in without interviewing anyone. You do not need to accuse anyone to catch it: the data shows it.
- Too fast: a 10-minute survey answered in 2.
- Straight-lining: the same score on every scale question or every row of a grid.
- Duplicates: two surveys with exactly the same answers.
- Odd timing: ten surveys in a row one minute apart.
- One interviewer whose results are very different from everyone else's in the same area.
SurvIQ flags the first three automatically in Responses for you to review. It never excludes anything on its own: you decide which response to remove from analysis, and you can include it again later.
5. If the survey runs on a tablet
At fairs, stores or events, the interviewer often hands a tablet to the respondent. Kiosk mode keeps the survey full screen and returns to the start after each response, so nobody sees the previous answers. Results can be filtered by channel, so you separate tablet responses from link responses.
A limit worth knowing
SurvIQ needs an internet connection to submit each response. If your fieldwork is in areas with no signal, you need a tool with offline mode, or plan for interviewers to submit where there is mobile data.
Daily control routine
- Mid-morning: progress by interviewer. Call anyone at zero.
- End of day: progress by area and quota. Reassign for the next day.
- End of day: review responses flagged as fast, straight-lined or duplicate, by interviewer.
- Once a week: compare results across interviewers in the same area.
If you run studies for clients:
SurvIQ for consultancies