Using the Internet, Social Media, and Mobile Apps: Northern Kentucky University, Greater Cincinnati Region

Research involving the use of social medial and the internet (i.e. Facebook, Twitter, MTurk etc.) has grown exponentially over the past few years.

It is imperative that the researcher be aware of the dynamic and evolving nature of utilizing these technologies in human subjects research.

Because this field is constantly evolving, it is impossible to identify every circumstance or type of research activity that may involve social media or the internet.

Social media is an internet-based mode of communication that allows users to interact with the medium (typically a website) and/or other users of the medium. Social media is a broad term that includes social networking (e.g. Facebook and Twitter), social photo and video sharing (e.g. Shutterfly and YouTube) and interactive websites.

Here is some guidance to help you create your protocol when utilizing the internet, social media, or mobile apps.

Data Mining or Passive Information Gathering

Data mining and passive information gathering often does not meet the federal definition of human subjects research because it lacks both interaction and intervention with the individual about whom the data is being collected. Examples of this would be public Twitter feeds, public Facebook profiles or wall postings, information from public/open chat rooms, etc.

If the individual or social media network site has not placed any restrictions on access to information, the researcher:

Can submit an IRB application for a formal, not human subjects, research determination.
Should ensure that all of the information on an individual is de-identified and that individuals not be individually identified and/or that the group cannot be readily ascertained.
Should note that the individuals on many public blogs use pseudonyms to conceal their identities. In general, researchers should avoid eliciting information from other sources to establish the real identity and exercise caution to ensure that they do not accidentally reveal a person's identity.
Should be sensitive to data that individuals may not realize is accessible (e.g. data left on directories that are accessible via use of a web crawler). Researchers should consider this data private unless they can demonstrate that data is sufficiently de-identified. If the researcher cannot demonstrate this, the researcher should submit an IRB application.

Deception Research

Deception research involves interaction with or intervention in the participant's environment and therefore meets the federal definition of human subjects research.

When using deception, researchers should:

Make sure individuals are aware that they are the subjects of an experiment. Totally blind deception experiments are not permitted.
Most often, individuals must agree to participate. It may be necessary to not reveal the true purpose of the experiment at the beginning, however, prompt and thorough debriefing is expected. Researchers should share the debriefing plan in the IRB protocol.
Unless the study involves minors, the researcher must utilize a plan to safeguard children from participating in research involving deception. These safeguards must be included in the IRB application.

Using Social Media to Recruit

Using the internet to recruit can help researchers cast a wide net when recruiting which can result in more reliable results. For the most part, the IRB will review internet recruitment the same way it would review in-person recruitment. Direct advertising for study subjects is the start of the informed consent and participant selection process. Researchers must submit an IRB application prior to using social media for recruitment. Please consider:

As with all human subjects research, the IRB must review anything a potential participant might see or hear related to a research project.
Consent and recruitment are separate activities and should not be combined. For example, recruitment/recruitment statements should be used before the consent process. It is generally not acceptable to consent the participant only as part of the recruitment message.
Researchers should ensure that recruitment of participants using social networking sites meet the criteria for equitable selection of participants and that the sample selection is justified.
Researchers must ensure safeguards are in place for screening children, prisoners and other vulnerable populations, unless these populations are the intended participants for the research study.

There are two types of social media recruitment:

1) Passive recruitment: Distributing recruitment materials (ads, posters, flyers) with the aim of attracting potential participants to contact the researchers for enrollment. For example, placing recruitment materials in patient support groups.

2) Active recruitment: Approaching and interacting with specific individuals with the aim of enrolling them in research. For example, emailing or “friending” a member of a patient support website based on their online activity and membership in the group.

If you are creating a Facebook page for the research project, the researcher should place the following information in the “About” tab in Facebook:

Purpose of the research
Purpose of the Facebook account
How the page will be monitored and who is responsible
Any limitations on how the page should be used
What can be seen publicly versus privately

Here is a great document explaining ethical use of social media with human subjects.

Use of Mobile Devices and Other Emerging Technologies

Researchers must submit an IRB application prior to using social media for recruitment. Please consider:

Consent and recruitment are separate activities and should not be combined. For example, recruitment/recruitment statements should be used before the consent process. It is generally not acceptable to consent the participant only as part of the recruitment message.
Researchers should ensure that recruitment of participants using social networking sites meet the criteria for equitable selection of participants and that the sample selection is justified.
Researchers must ensure safeguards are in place for screening children, prisoners and other vulnerable populations, unless these populations are the intended participants for the research study.

The Food and Drug Administration (FDA) regulates a subset of mobile apps which meet the definition of “mobile medical app” (MMA). The FDA has stated that the agency intends to exercise enforcement discretion with respect to mobile apps that may meet the definition of medical device but pose a low risk to participants.

High risk MMA examples –

An app that monitors blood pressure and transmits the data to a treating physician.
An app that controls the delivery of insulin through an insulin pump.

Low risk MMA example –

Apps for recording patient diet and exercise habits.

There are five core areas to understand before designing research on or with mobile apps:

General Considerations and Risks

Some of these common and still pertinent risks include:

What data will be collected?
- Research data?
- Incidental data captured by the app?
Is data collected identified or de-identified?
Where will the data be sent? (Medical records?)
What are the risks to privacy and confidentiality?
What are the risks of the app not working as intended?
Are there costs to the participant for using the app?
Is the app content (font, language, videos) appropriate for the participant population?

Technology

If the app being studied, or used in a study, is one which is created expressly for that research, since the app is custom built, extra precautions including adhering to best practices in software development should be in place. Commercially built apps remove the development burden from the researchers, however they present their own concerns.

Some high level concerns for technology include:

How will the researchers access the data?
Will the devices be provided or will the participants have their own?
Do participants use their own logins and accounts?
What types of devices will be supported?
How will software updates, new phones, and study cancellation be handled?
Should the app have the ability to be remotely reset or stopped?
Are the researchers now required to review data in real time from the remaining participants?
If the app is disabled remotely, how will that be communicated to the researchers?

Data Security and Confidentiality

Considerations include:

What mechanisms are in place to minimize risks to confidentiality?
Where is the data stored?
Is it encrypted?
What passwords or pins are used for the participants and investigators?
What auditing is available on who accessed the data?
How is access to third parties evaluated and restricted?
Is the device itself collecting data?

Consent and Terms of Use (TOU)

How is consent obtained?

1. eConsent?

Is it combined with the mobile app?
Does it contain all elements of informed consent?
Is there an option to use written paper consent instead?

2. Waived consent?

3. Written paper consent only?

In addition to consent, there may be legal terms related to the technology (terms of use, license agreements, privacy policies, etc.) that the participant must accept in order to use the app.

Questions include:

How is the information explained to the users? (These agreements are often just clicked through.)
Does the content comply with applicable research regulations?
Is there any information in the agreement that should also be included in the consent form (e.g., access to data)?

Support and Training

Researchers should consider:

How will participants get support for the technology, both at study start and ongoing?
How will password resets be handled?
Is there any monitoring in place to ensure the technology is working?
Who is providing all of this support? What hours?

Information presented by Advarra, February 21, 2018. https://www.advarra.com/resource-library/mobile-apps-considerations-for-use-in-research-involving-human-subjects/

Amazon Mechanical Turk (mTurk)

What is MTurk?

MTURK is a website run by Amazon that works as a readily available marketplace to match "workers" with available work from various "requesters". Amazon describes MTURK as, "a marketplace for work that requires human intelligence. The mechanical Turk service gives businesses access to a diverse, on-demand, scalable workforce and gives workers a selection of thousands of tasks to complete whenever it's convenient".

MTurk allows researchers to have access to a large population of willing participants for research studies. Researchers can generate a HIT (Human Intelligence Task) that gives MTurk users a title and description of the online task (recruitment announcement). The HIT includes:

The title
Description of the online task
Amount of time it takes to complete the task
Compensation

What do academic researchers need to consider when submitting a MTurk study to the IRB?

Recruitment: The recruitment announcement (HIT) must be submitted to the IRB for review. Researchers should:

Explicitly state that the study is "research" and not a "job”. Researchers should be clear about compensation and bonuses.
Be clear about how long it will take the participant to complete the task.
State how long it takes for the Turkers to receive payment.
Address whether or not the compensation is contingent upon certain conditions.
Ensure that the complexity of the task and the amount of time expected for completion is reasonable and communicated clearly.
Tell participants if there is a screener in order to qualify and if the Turker gets paid to complete the screener.
Tell participants if any extraneous software is required to complete the tasks; this should be stated in the description.
Describe the type of task (survey, watching videos, etc.).
State the researcher's name and/or school affiliation. This should be listed either as the Requester or in the description of the HIT.

Consent: The first page of the online survey should include the consent document. Researchers will request a waiver of documentation of informed consent in Mentor IRB when submitting their IRB application. Consent will be verified by selecting "I agree" at the bottom of the survey page. If the participant selects "I do not agree" they will be taken to a page thanking them for their time. The NKU informed consent template has language specifically related to MTurk studies.

Debrief: If deception or incomplete disclosure is being used, the last page of the survey could include the debriefing information.

Confidentiality: MTurk for research purposes is considered confidential, not anonymous. Recent research shows that MTurk worker IDs can easily be linked to an individual’s Amazon profiles including the individual’s wish lists and previous product reviews. This means researchers must use caution in deciding what information to collect from participants. The default should be that participant MTurk worker IDs not be collected. If it is necessary to collect worker IDs, then the researcher should ensure that worker IDs are kept confidential and secure, are not linked by to study data and are deleted after use.

Researchers should be familiar with the Amazon MTurk (AMT) Terms of Service (TOS) as well as other online survey softer ware TOS if used. TOS especially as it related to the collection of participants online behavior and history (through the use of cookies or other tracking systems) and the selling of participants data to third parties.

Excerpted from https://www.umass.edu/research/guidance/mturk-guidance

Resources

https://www.umass.edu/research/guidance/mturk-guidance

https://researchservices.cornell.edu/compliance/human-research

https://www.quorumreview.com/wp-content/uploads/2012/12/IRB-Review-of-the-Use-of-Social-Media-in-Research_Gearhart_Quorum-Review_Monitor_2012_12_01.pdf

Source 2015: Quorum Review IRB presentation by Dominic A. Chiarelli, JD, CIP Regulatory Attorney

https://www.quorumreview.com/wp-content/uploads/2011/12/UsingSocialMediainResearch.pdf