Web-based studies

Contents
  1. Informed Consent in Web-Based Studies
  2. How do participants indicate consent?
  3. Anonymity of online participants
  4. Compensation for web participants
  5. Withdraw or drop-out in web experiments
  6. How to conduct your study
  7. Recording interviews online
  8. Additional resources

Informed Consent in Web-Based Studies

Requirements for obtaining informed consent hold whether data are collected in person or over the web: In order for a potential subject to make an informed decision as to their participation in any research, the subject must be made aware of the nature, purpose and duration of the research, the method by which the research is to be conducted, and any risks or benefits which may come from participation in the research.

The information provided to the subject should be written at an appropriate level and should avoid the use of jargon or technical terminology.

For a summary of information which should be included in consent forms more generally, please see this page. When you implement your study, you can use one of our web-based study consent templates, available here. You can either directly display this or provide a link to your form, as in the screen shot below. If you choose the latter, please summarize what participants are consenting to on that page. For example, if you are asking for consent for general public use of data, then put that in the bullets for both in your consent form and in the bullets where the link is displayed.

ExampleWebStudyFirstPage

How do participants indicate consent?

  • Other LEL Ethics consent templates contain a place for the subject to indicate their consent by a signature. For web-based studies, you may replace the signature area with another method by which subjects indicate their consent.
  • For example, you may instruct subjects that by making a particular response (e.g., choosing one of two radio buttons, or clicking a ‘continue’ or ‘accept’ button), they are consenting to participate in your study.

Anonymity of online participants

Ensuring anonymity can be complex when data are collected online, particularly with a third party company is involved. Please be aware of the following:

  • Participants IDs provided through web-based recruitment services (e.g., workerIds on Amazon Mechanical Turk) are effectively linkable to personal information.
  • Websites like Amazon Turk, Survey Monkey, and Qualtrics store participants IP addresses by default.
  • You may need to anonymize data before sharing it, publishing it, making it public, etc.
  • At the same time, you may not be able to prevent third-party companies like Survey Monkey, AMT, or Qualtrics from storing data of some kind. Check the terms of service to find out (or click links)
  • You can find out whether a third party is part of the Privacy Shield agreement (a legal framework agreement between EU and US to ensure compliance with the Data Protection Act).

For these reasons, some general disclaimers may be in order:

“Although every reasonable effort has been taken, confidentiality during actual internet communication procedures cannot be guaranteed.”
 
“Anonymous and/or aggregated data collected for this project may be stored and used by Qualtrics.”
 
“Any data we share with other researchers will not include information which could identify you (e.g., IP address or identification number).”
 
For interviews conducted over email:
“Email is an unsafe form of communication for private responses. This is because email can be easily hacked. Therefore you should only take part in the study if you/your company are prepared for your responses to be made public, even though the research write -up will not link any responses to individuals/companies.”
 

Please take note of all Terms of Service agreements for these services. Here is a selection of the types of “HITs” that violate Amazon Mechanical Turk ToS:

  • HITs requiring disclosure of the Worker’s identity or email address, either directly or indirectly (modulo worker ID’s…)
  • HITs requiring registration at another website or group
  • HITs that have explicit or offensive content, for instance, nudity, but do not have the Adult Content Qualification
  • HITs asking Workers to solicit third parties
  • HITs that require Workers to download software

Compensation for web participants

Check out this page about fair compensation for web participants.

If you are using a platform (like Crowdflower) that takes a cut of participant payments, or involves a third-party site which takes a cut of payments or provides site ‘credit’ in lieu payment, you must be upfront about this.

Indicate the base-rate you will pay, but state explicitly that compensation may vary. For example,

“We will pay £5 to have you participate in this study, however please note that your actual compensation may vary depending on the terms and conditions of the site you are accessing this study from.”

Please note: ethical research practice dictates that participants must receive the stated compensation — that is, you cannot reject a participants’ work.

Withdraw or drop-out in web experiments

Participants must always be allowed to terminate their participation in a study.

In lab-based experiments, participants are often given the option of withdrawing from the study at any time. That means not just during the study, but any time after their data has been collected. For web-based studies, this may be more complicated. Please consider:

  • If you cannot collect information allowing you to connect the participant to their data, then you may not be able to delete a participant’s data.
  • If participants terminate their participant mid-experiment, you may have collected partial data.
  • Or, you may have recorded the fact that they terminated the session.

Can you keep this data?

You should inform participants during the consent process if you will maintain partial data. Recording only information about how many participants terminate your experiment (i.e., without recording any other data) may not require any special statement.

  • By default (see consent template) participants should contact the researcher, and request that their data be deleted.
  • If this is not possible, there are some alternative options. For example:

“Your participation is voluntary and you may stop at any time and for any reason. If you stop participating at any point, data collection will end, but data already collected will be retained.”

“Your participation is voluntary and you may stop at any time and for any reason. If you would like all your data deleted, click the ‘exit’ option at any point during the study. Otherwise, data already collected will be retained.”

If you encounter other ethical issues in collecting web-based data, please contact lel.ethics@ed.ac.uk. 

How to conduct your study

There are a number of services available to facilitate conducting studies over the web. Please be aware of whether, when collecting data in this way, a third partly may have access to your data (permanently or temporarily).

  • Amazon Mechanical Turk: Amazon will collect and store some data, even if you are collecting and storing most data on a University-approved server.
  • Prolific Academic: similar to Amazon Mechanical Turk (see here for more information about GDPR-compliance)
  • Qualtrics: University-approved service, all students and staff can access via this link.
    • JISC Online Surveys: University-approved service, with access granted via Information Services (IS.Helpline@ed.ac.uk) and students will need their supervisor to confirm suitability of use

Recording interviews online

It is possible to record audio over the web. You can do this either via a website linked to a University (or other GDPR-compliant server), or using another method. Please consider the following issues around data security:

  • If you record a Skype interview, then Skype retain the recording for 30 days on their server.
  • Zoom do not store the recordings and instead these are saved to local device or Zoom cloud (for paying customers) (Archibald et al 2019) – but not these do not appear to be supported on IOS or android
  • You could record using a secure recording device you have in person, rather than using one of these services.

Additional resources