Safeguarding, safety and wellbeing in user research

Neontribe recently ran some workshops on the topic of user research for Tech for Good Build participants. In the workshops, we talked a bit about how we often need to reflect on the safeguarding challenges around conducting user research with vulnerable users. Many of the themes that we wrote about in this blog early last year still ring true today and we thought it'd be worth sharing again.‍

What should you do when you ‘make’ your user research participants cry?

I’m spending a lot of time thinking about this at the moment. Along with a whole set of follow up questions:

• How do we design our research and participant selection in ways that might help us predict and be prepared for when people might become upset?
• How do we respond in the moment, to create the best experience for the participant, continuing the session if that is important to them?
• How do we provide support to our researchers so that situations like this don’t take a toll on their emotional wellbeing?

So far I’m certain of only two things.

1. It isn’t necessarily bad that a participant cries.
2. The researcher probably did not actually ‘make’ the participant cry, but will often feel like they did.

Neontribe and user research

Neontribe is a digital agency that spends half our time working in the charity sector. We’ve always been huge advocates of user research, with most of our developers as well as our UX/UI and content/strategy folk participating in sessions. When the company started, we delivered a couple of projects with Channel 4’s youth engagement team so we’ve always had some understanding of safeguarding. Our approach has been to ‘follow the safeguarding procedures of our client organisation or charity’. Broadly we think that’s a good start.

But over the past couple of years, we’ve encountered some situations that have made us want to do more.

Researching abuse is emotionally challenging

In 2017–18 we worked with The Haven, Wolverhampton as part of Comic Relief’s Tech vs Abuse programme.

We knew the project might be distressing, and it was. Even at the horizon scanning stage, our developers found that testing existing tools designed for people experiencing abuse led them to soul search about the nature of their own relationships. In usability testing we needed to provide support after a participant told us “its ok, I’m only crying cos that one [of the women’s stories we had used] could have been me”. And we had several participants who wanted to tell us about their journey out of abuse, before they felt ready to engage with the testing process.

We followed good practice

We’d taken some steps in advance, broadly following The Haven’s safeguarding policies.

• We always had Haven staff present to provide support and cover safeguarding responsibilities, even when it meant they had to find volunteers to cover staff illnesses, or rely on external colleagues they’d trained.
• We researched in pairs, to protect both participants and researchers.
• We planned extra time for individual sessions, because we knew people might want to talk about their experiences as well as engage with our tasks.

We could have done more

But we hadn’t done other things that would have been useful, both for us and the client. We hadn’t:

• Thought about how to create time, within working hours, for researchers to process the work’s emotional impact.
• Thought about whether we could, or should, be trained in responding to disclosures, or whether to carry out enhanced DBS checks on our staff. That way charities could rely more on our staff, rather than always needing their staff to be present.

Talking about safeguarding processes is also difficult

The second situation was one where we were less prepared.

We were doing user research for the National Council for Voluntary Organisations, helping them and their partners create new web pages explaining safeguarding responsibilities to charities working less regularly with people at risk.

The research participants were all people working or volunteering in charities. We had exercises designed to help us create process focused user journeys such as “how do you find information about your responsibilities” and “how do you share it with colleagues”. We had not foreseen a need to plan for participant distress.

At two sessions, participants arrived already visibly stressed, leading us to quickly decide that we couldn’t ignore their immediate needs. Instead we listened, whilst they vented, almost tearfully, about the strain of implementing safeguarding practices in their organisations. But afterwards, we second guessed ourselves. If we had stuck to the original research plan, would that have been less distressing for them?

When we think about our research participants only through their professional identity it becomes easy to forget their human identity underneath. Safeguarding professionals have a saying that “no-one should ever deliver safeguarding training without being prepared to handle disclosures and other difficult conversations”. Looking back we could have applied this saying to our user research.

So what are we doing?

We’re:

1. Buying in a “Mental Health and Wellbeing’ training day for Neontribe staff.
2. Reflecting on a pair of articles about research produced by Snook. One covers emotional well-being of participants and the other considers looking after yourself.
3. Writing a set of safeguarding procedures for user research. They will include checklists for planning sessions and recruiting participants.
4. Exploring the need for, and legal issues around, enhanced DBS checks for researchers on certain projects.
5. Questioning whether the flexibility of our research sessions (we’re very free-form) is more problem or strength when it comes to supporting participants.
6. We’re a flat structure company with no line management so can’t use that system as a safety valve. We’re looking at what other forms of 1-to-1 relationships might support staff. For example could the supervision approach used within counselling or volunteer management be useful?
7. We’re exploring the Holmes and Rahe resilience measure, introduced to us by this Home Office Digital blog about psychological safety for researchers.

‍

Additional resource: DigiSafe is a step-by-step digital safeguarding guide, designed for charities designing new services or taking existing ones online.

With thanks to Joe Roberson.