How to get consent via SMS in phone interviews

This article is accompanied by this sample form which illustrates what is proposed below.

Informed consent is a cornerstone of ethical research, and an absolute requirement in a lot of the work done by SurveyCTO users. For low-stakes enquiries, sometimes verbal consent is acceptable, and this could also be verified using an audio audit recording. Other face-to-face studies have used the image field type to capture the signature of the respondent, noting their consent. However, at the time of writing this article (May 2020), as we all move to adapt to the reality of COVID-19, we'll need new methods for interviewing, and for recording consent.

Computer-assisted telephone interviewing (CATI) is one alternative data collection method, but what about consent? Will verbal consent be good enough? If yes, great, but given that Google prevents call recording after Android version 7 for security reasons, reliable call recording becomes harder to set up, but it is achievable.

What can you do? What is an easy, none too difficult, verifiable method for recording consent?

Requesting consent via SMS

You can request consent via SMS in two ways. Thanks to the extensible nature of SurveyCTO through field plug-ins, you can make use of the launch-sms field plug-in to send an SMS directly from a form. The message field in the SMS app that is launched can be pre-populated with the consent statement, as well as response options the respondent can send back.

The above field-plug-in strategy does have limitations and challenges though. Instead, you might use a bulk SMS service to send hundreds, or even thousands of SMS's to respondents at once. This can be an alternative or complement to using the launch-sms field plug-in. Please read this and the above linked article for a full picture.

Using the launch-sms field plug-in to request consent

The whole process might work as follows:

  1. Call the respondent (for example, using the phone-call field plug-in. See the CATI starter kit for more details).
  2. Ask to speak to the right person.
  3. If the respondent is available, let them know how the informed consent process works. While the enumerator could read out a statement, we recommend including the informed consent statement in the SMS too, for the most ethical treatment of respondents.
  4. Use the launch-sms field plug-in to send an SMS. The SMS should include instructions on how to respond (e.g. "Reply YES if you agree"). This can happen while still on the call.
  5. The respondent opens the SMS and replies with the appropriate confirmation. This step can also happen while still on the phone, but some respondents might be less comfortable multitasking on their phones, so they might need to be called back.
  6. When the SMS is received by the enumerator, they can open the SMS app, and take a screenshot of the response. The screenshot should display a.) the phone number of the respondents (as proof of who it came from), and b.) their response ("yes").
  7. With the screenshot created, the enumerator can then use an image field to pick the screenshot from the gallery, attaching it to the blank form. To avoid confusion and for privacy reasons, enumerators should delete screenshots between interviews.
  8. With consent on record, the interview can continue. 

Following the above procedure, you'll have an auditable record of informed consent being granted.

There's also another accessible record of SMS exchanges: the device's storage. The Google Play store has many apps that can export the SMS inbox of an Android device to an accessible file. Perhaps that file might also be backed up to a location you have access to.

Limitations and challenges

The above isn't a perfect solution, but its strength is that it doesn't take much to set up, and it can work. For less well-resourced projects, this will be the best and fastest solution to implement. To help you make an informed choice, please do keep these drawbacks in mind:

  • This process can take a lot of time collectively, since several minutes will be spent on dozens or even hundreds of respondents. This can add up to several worker-hours just sending and receiving SMS text messages.
  • Some respondents may not have enough phone credit to send a response.
  • Some respondents won't be invested enough in what you're asking them to do to spend their credit on an SMS.
  • This process can be awkward. The multitasking required in steps 5 and 6 above might be difficult for some respondents, so the call may need to end to allow them to respond to rely on the SMS.
  • While the above might be 100% necessary for your study, collecting consent in this way is a point of friction, and therefore, a point where you may lose the respondent.

Alternatives and enhancements

While the above isn't a perfect solution, if you have to collect data which requires a record of consent, and if you can't pursue face-to-face data collection, you'll have to make a plan. Here are some alternatives and some enhancements to consider. 

  • Incentives for respondents can help get them through points of friction. See the "2.4 Third party services (optional)" heading in the CATI starter kit for more. Incentives generally can help to reduce friction, and increase response rates.
  • Skip sending SMS's for consent by investing in an enterprise-grade solution for call recording to eliminate friction. For example, J-PAL South Asia has built an integration with the Exotel cloud phone solution, which facilitates recording.
  • Send SMS's requesting consent in bulk, as described above.

Do you have thoughts on this support article? We'd love to hear them! Feel free to fill out this feedback form.

0 Comments

Please sign in to leave a comment.