New View of Consent
In the course of Public Policy Lab’s work, we conduct research and co-design with members of the public, learning from them in order to fuel our work and improve public systems. This kind of research always runs the risk of being extractive, of allowing ‘experts’ to sample the life experiences of ‘respondents’ and spin up a shiny deliverable, rather than generating results that speak to the needs and preferences of participants.
So, as ethical practitioners, we look for ways to mitigate power inequality in our research relationships. One key tool is our process for seeking informed participant consent. Our goal is to empower individuals to understand the benefits (and harms) of participating in research with us and to enable them to exert control over what we can (and can’t) do with the information they share.
This post is about how we first realized our consent processes were insufficient — and how we’ve upped our game since then. It also describes seven questions that inform our consent practices:
- Are you offering participants fair compensation for their time?
- Are you conducting the consent process in plain language?
- Are you maximizing participants’ control over their data?
- Have you made it clear that the research is not confidential?
- Are you collecting as little personally identifiable information as possible?
- Have you been explicit about potential harms?
- Are you prepared to provide resources if people are having problems?
We hope all design researchers will ask at least these questions of their own processes for seeking informed consent. Read on for how we arrived at these key questions and where we still want to go from here.
Rethinking Consent
A major turning point in our consent practices came in 2015, when we conducted research on jail overcrowding in Louisville, KY. (NB: This piece builds on our previous post, Good Intent is Not Enough. We’d recommend reading that piece first so you have context for this story.)
Quick Thinking Required
The first time that a few members of PPL’s team visited our government partners in Louisville, we also went to see the city’s main jail, expecting only to tour the facility and meet its leadership staff. During our walk-through, the jail director surprised us with a question that set in motion our forever project around consent:
“‘Well, do you want to meet with some people?”
We wanted to, but we were not prepared — creating the research plan was something we’d planned to do when we got back to New York! Standing in a cinder-block meeting room sandwiched between the jail’s dorms, the team had a quick huddle. We weren’t certain the people we’d be speaking with could meaningfully consent to participate in research since they were not in jail voluntarily. We also didn’t know the legal ramifications of collecting data about participants’ experiences. Could PPL’s research notes be subpoenaed?
Jail staff suggested we could use the jail’s standard consent form, but when the team looked it over, we saw it didn’t describe jailed research participants’ risks or protect their rights. Like most commercial consent forms, this document’s job was to say, ‘Whatever you say today, you can’t get mad at us later.” (Honestly, it wasn’t even that much worse than PPL’s standard consent form at the time.)
Because we were not confident that we could appropriately seek consent from the jailed people with whom we’d be working, we created a quick workaround: We would ask them to sign the jail’s consent form, but also request that they not share any information specific to their personal stories or charges.
The jail agreed to let us work with the jailed people without staff present. After staff brought several jailed people to the meeting room, our design fellow Jen Thibault and director Chelsea Mauldin explained the project. We asked if they’d help us understand what would happen if one of our design team got arrested in Louisville.
“What have I been arrested for?” Chelsea asked the participants.
“For public intoxication!” they replied.
“Of course,” Chelsea said. “What happens to me next?”
From there, participants walked Chelsea through the process of being brought to the jail, booked into the facility, going to court, and ultimately being released. Jen captured Chelsea’s arrest story on a whiteboard on one wall of the meeting room. By using role-play, we were able to create space for participants to share their knowledge of these processes without asking them to disclose personal information.
First Steps Toward a Participant-Centered Consent
Back in New York, we knew we had work to do to prepare for our next trip to Louisville, when our whole research team would be present and we’d certainly be returning to the jail. We’d asked PPL research fellow Kate Krontiris to join the project specifically because she had prior experience working with the criminal-justice system. She led the team in developing a new consent process designed to project and empower our jailed participants.
First, we researched getting a Certificate of Confidentiality from the National Institutes of Health (NIH) to protect identifiable research information from forced or compelled disclosure, but we did not meet the eligibility criteria. This meant we could not guarantee confidentiality to participants in the (admittedly unlikely) event that the jail, a prosecutor, or other law enforcement subpoenaed our notes.
As a result, we maintained the practice of asking our participants to not disclose anything they didn’t want this jail or prosecutors to know — and avoiding lines of questioning that dealt with crime. We also strengthened the part of our consent form about disclosing the potential harms that a participant might experience from engaging with us. We wanted to do everything we could to make sure they knew what they were signing up for.
This included telling participants that they didn’t have to participate at all. It was impossible for us to be sure that our ‘volunteer’ participants truly wanted to be there with us, so we told them we’d be happy to just hang out in the interview room with us for their allotted time, talking about sports or pets or the weather. The jail didn’t have to know if they answered our questions or not.
In the end we interviewed 14 jailed people and 13 members of the jail staff. They generously provided us with invaluable information about conditions and needs inside the jail and in the community. We tried to repay their trust by collecting that information without any personally identifiable information — and by using it to advocate for changes in the jail system.
Seven Questions for Informed Consent in Design Research
Since 2015, we’ve redesigned that original Louisville consent process at least a dozen times. We’ve come to understand the act of seeking consent as being at the core of our professional practice. As we wrote in our previous post, “consent offers researchers, designers, and policymakers a real-life tool to center agency, dignity, and honesty in the policy-design process.”
Consent for clinical research — for human trials of a new vaccine, for example — generally includes these key elements:
- a statement that the project involves research and participation is voluntary;
- a summary of the research project, including goals, length of participation, and expected activities; and
- the risks and benefits of participation;
- an explanation of confidentiality; and
- alternatives to participation or an opportunity to withdraw.
PPL follows these same principles in our consent processes. While our design-research plans do not always require review by an Institutional Review Board (IRB), we believe that design research should always strive for the highest ethical standards, even when it’s not subject to the same institutional scrutiny.
In addition, we’ve also developed a set of key values and questions around consent that speak specifically to the type of research we conduct (qualitative, driven by human stories) and who we often conduct it with (members of marginalized communities who have good reason to be suspicious of researchers). We ask these seven questions of our research teams — and we think you should, too.
1. Are you offering participants fair compensation for their time?
Time is valuable, especially for our lower-income participants, who may be missing an opportunity for wages by taking the time to speak with us. And the task that we’re asking them to undertake with us — sharing the stuff of their lives, their personal stories, sometimes traumatic ones, so that our work can benefit — it’s not trivial or always easy. The least we can do, at a material level, is to offer compensation for their time.
We traditionally have provided research stipends to members of the public in the form of physical gift cards, often for a local drugstore chain. We provide the gift card at the beginning of the conversation, to avoid any suggestion that the stipend may be withheld until we can gauge the participant’s ‘performance.’
In shifting to remote research in 2020, we began to use a digital platform to buy, track, and manage incentives. Because rewards come in the form of gift cards, Visa cards, or cash, participants are able to exercise more choice in how they’ll be recognized for their time. Typically we offer $15 rewards for engagements of less than half an hour, and provide $25 for engagements between thirty minutes and one hour.
PPL is occasionally reimbursed for participant stipends by our project partners, but when we don’t receive reimbursement, we still provide the compensation. It’s a cost of doing our jobs well.
2. Are you conducting the consent process in plain language?
More than 30% of adults in America have low-literacy skills in English. To make our consent form easy to understand, it’s written in a question-and-answer format and at an eighth-grade reading level. (We ask our researchers to use a Flesch-Kincaid reading scale tool — built-in to Microsoft Word and also available online — to verify that the reading level is simple enough.)
Plain language won’t help people who have significant literacy or vision challenges, however. And people have pride — they generally don’t like to share when it’s hard for them to read. That’s why PPL researchers also verbally step participants through each section of the consent form. When we expect that participants in an upcoming research session may not speak English as their primary language, we seek to bring translated consent materials and/or a researcher who can translate the consent process.
When conducting research remotely, our researchers email a digital-survey version of the consent form to participants, but they are clear that the form does not need to be filled out in advance of our session. The beginning of each research session is dedicated to speaking through the consent form and double-checking that those who have already filled it out have understood what they consented to.
3. Are you maximizing participants’ control over their data?
Rather than asking people to wholly participate in our process or not, we ask participants to opt in to different ways of their story being documented.
Through a series of yes/no questions, participants tell us if we can take notes, audio record, directly quote them, take photographs and video where their faces cannot be seen, and/or take photographs and videos where their faces can be seen. This granular approach gives participants the ability to ‘toggle’ on or off gradations of consent as we work together.
At PPL, we also think that our participants have the right to change their minds. We clearly state that they are welcome to stop the session at any time, for any reason. At the end of each session, we allocate a little time to give the participant the opportunity to review what they had consented to and whether they wish to withdraw any of their data.
We also show respondents the photos that we’ve taken of them and ask them if there are any photos they want us to delete. In remote engagements, this includes screenshots. No one wants to be forever captured as the person with a funny bit of hair sticking up!
After all that, we leave participants with our contact information, so that they can get in touch to revoke their consent after even we’ve gone, if they so choose.
4. Have you made clear that the research is not confidential?
A primary privacy challenge in our job as qualitative researchers is that we are collecting data about individuals’ personal stories — and then sharing them with people who run the systems those individuals use or work in. We often cannot maintain the rich humanity of a participant’s experiences if we totally strip a story of its context. As long as we retain elements of context, there is a slight possibility that we may make participants recognizable to individuals who know them personally.
Further, the pool of potential respondents we’re speaking with may be known to people in authority, because they may have given us permission to conduct research at certain sites. Therefore, we make a point of explaining that we cannot promise total confidentiality, meaning that no one will ever know that an individual may have participated in our research.
We’ve run into situations where a partner’s legal team (perhaps more used to human research in the form of totally de-identified quantitative surveys) wants our consent form to promise confidentiality. But knowing that, as an organization, PPL can’t guarantee this level of total privacy gives us clarity in these conversations and the ability to direct partners toward the ways we avoid collecting and sharing personally identifiable information.
5. Are you collecting as little personally identifiable information as possible?
Personally identifiable information (PII) is the term researchers (and lawyers) use for names, dates of birth, Social Security numbers, addresses, and other data points that make a person uniquely identifiable. We do everything in our power to minimize the personally identifiable information we collect about our participants, while learning and knowing as much as possible about their stories.
We avoid asking for or documenting PII, and when we’re converting raw research data — notes, recordings, transcripts, text responses — into distilled quotes, stories, or insights, we try to remove identifying information that a respondent may have included. The less PII we collect, the less we have to protect, and the more confident we can be that our respondents will remain anonymous, meaning that their identity will not be explicitly linked to our products.
Whether working remotely or in-person, we store our consent forms — which are the only place where we collect a participant’s name and email or phone number — separately from the stories and materials that we collect from them. This means paper consent forms are stored in a lockbox away from project content, and digital consent is collected in an application that’s separate from the tools we use for storing project data. Instead of linking data points to a name, we have a protocol for creating a unique code for each participant. We use this code on everything from the interview notes and stimuli used in research sessions to our data repository.
6. Have you been explicit about potential harms?
When seeking consent from participants, we explicitly disclose the possible harms of participating in a research project with us. Usually these harms are limited to the potential for someone to know that a respondent spoke with us (see above), but we’ve also described, for example, how our research could trigger unwelcome memories or emotions (as in our mental health research) or create legal exposure (as in our jail project, above).
Similarly, we disclose that our commitment to privacy is not absolute: if a research participant tells us that they are being hurt or that they are planning to hurt someone else, we will try to notify someone who can help. While we are not mandated reporters (professionals who are legally required to file a report when abuse is observed or suspected), we are human. We have an ethical duty to intervene if we think someone is in danger.
(Frankly, this topic could be its own blog post: the circumstances that led us to this practice, the effects on our researchers of parsing other people’s trauma, PPL’s internal reporting processes, the challenges of finding a safe and appropriate person to disclose to, etc…)
7. Are you prepared to provide resources if people are having problems?
While reports of ongoing danger are thankfully rare, we do often hear painful stories of the human cost of service-delivery problems. As outside-the-system researchers, it’s generally not appropriate or effective for us to intervene or assist with individual social-service cases. But while we may not be able to help personally, we do come prepared with contact info for someone who can.
We design our hard-copy consent form in a way that allows participants to keep the first page. This page explains the project, what the participant will be asked to do, and provides information about their rights. The document also includes contact details for resources like mental health crisis lines, an agency ombudsperson, or other providers who can help if a participant is struggling.
In the course of conducting research, if we hear a respondent describe challenges for which we know they could get help, we remind them at the end of the session that the form has helpful contacts. Of course, we also include contact information for our team, so that participants can reach out to ask us questions later.
Continuing This ‘Forever Project’
The work of improving our processes for informed participation is far from done. We’re still not satisfied with our methods for sharing the products of our research with participants, for example, and our pandemic shift to remote collaboration has created new questions around access and privacy that are not yet resolved.
More fundamentally, the process of research remains unequal: we, the professional research and design team, will go off and make things informed by participants’ stories. For those things, we will earn money and have access to power that exceeds what we currently offer to participants. We’re continuing to explore how to better distribute the value our research generates.
Our minimum responsibility is to ensure that our research participants can make meaningful and informed choices to share their stories and selves. We make every effort to reduce harm, but we also respect the right of our adult respondents to choose disclosure. Some people want their stories heard. They are proud to share their face, their voice, and the details of their experience when they think it will be of value. We can support that openness just as much as we seek to protect others’ desire for anonymity.
In any case, once a participant has said how they want to participate in our research — given the clearly-stated risks and rewards — then it’s our job to fulfill the trust that they have placed in us.
