Informed Consent

What are the challenges and strategies for obtaining informed consent in technology research involving people with dementia?

2024-2026
Part of PhD programme at Eindhoven University of Technology
Activities: semi-structured interviews, thematic analysis
Published at CSCW 2026:
Ruitenburg, Y., Lee, M., IJsselsteijn, W., Markopoulos, P. (April, 2026). Beyond the Signature: Addressing Informed Consent Challenges in Dementia Technology Research. In Proceedings of the ACM on Human-Computer Interaction (CSCW007), 10(2), 1-29. https://doi.org/10.1145/3788043

The moment that started this research

In the first year of my PhD, I did an interview study involving one-on-one interviews with people living with dementia. I thought I had prepared everything correctly. I had gotten ethical approval from the university. I had asked the formal caregiver who would be willing to talk to me. She had given proxy consent for me to talk to this person. I had redesigned the consent form to be simplified and easier to understand. And I had taken the participant through every single element, and he had said yes to all of them.

But as I started the interview, I got up, walked to the camera, and turned it on. He said, "Oh, is that also part of it?" I said, "Yes, so I can record the interview and remember what you said later on." He said, "Oh, I don't like that." So I turned off the camera and decided to just use audio recording instead.

I asked my first few questions of the interview, but he was very hesitant in his responses and did not look very comfortable. So I asked him whether he wanted to be here with me or would prefer to leave. Still, I could not interpret most of what he was saying. So after two minutes, I decided to end the interview and return him to the group.

What confused me was that I had followed every ethical procedure. So how did I still end up with a participant who did not want to be there?

And note what set off this experience. It was when I turned on the camera, so when I started using the technology. While he had given consent on paper, in the room, in reality, it was a different story. And this must happen more, all over technology research, as we as design researchers are involving technologies for which we do not quite know what the impact will be on the participant.

Informed consent and dementia

I think we all know what informed consent is. It is there to respect participants' wishes about what will or will not happen to them in a study. It relies on information, comprehension, voluntariness and legal capacity.

When we do this with people with dementia, all of these elements become difficult, because people with dementia can experience memory difficulties, communication difficulties, and challenges in understanding abstract concepts, such as the risk of a study. Some of them may also lose the capacity to give consent themselves.

What we usually do in participatory design with people with dementia is process consent. This is a five-step process. First, we ask a formal caregiver or a family member for permission to approach the participant. Then we check whether the participant has the ability to understand the study and the risks it may involve. Then we obtain their initial consent, either verbally or through a signature. Then we continue to monitor their ongoing consent during the study. And afterwards, we support their return to their group or their normal lives.

While this is theoretically a really good model, the literature did not describe whether it actually worked in practice. Talking to my colleagues, and experiencing the situation I described myself, made me realise that it does not really work that well yet.

This fits within larger discussions in Human-Computer Interaction (HCI), Computer Supported Collaborative Work (CSCW) and design research, where people are saying that the informed consent process is too long and complex, that it has become too rigid, and that it is too formal to fit with our exploratory and participatory way of working.

What we did

I held an interview study with 3 junior researchers (PhD candidates), 3 senior researchers, 2 data stewards, 3 people with dementia, and one partner of a participant with dementia, who all had experience with the informed consent process in dementia technology research. I asked them: what challenges do you face, and what strategies do you use to address them? We used reflexive thematic analysis to analyse the results.

Four challenges

What we found was four main challenges that still exist despite the process consent model.

1. Conveying information without overload.

A participant with dementia said, "If I read that [last sentence], I have already lost the first". Because comprehension and memory are affected, the whole letter already feels overwhelming. Another participant said, "Pfft, how am I supposed to [...] fill all that in?" The study becomes burdensome before the study has even taken place. And this makes researchers wonder: "Did they actually mean yes [...] [or] did they say yes because they like you?" They wonder whether they actually got the consent they needed.

2. Documenting agreements without frightening participants.

As a senior researcher said, these forms make it feel like you are "getting into a contract". The participants with dementia also checked: they would read every single line and ask, “are there snakes in the grass?” A junior researcher said that the formality and rigidity of the process make the study “seem way bigger than what
you are actually asking of people". This can add strain for the researcher and fright to the study, which of course we do not want in this type of research, because we want to observe natural interactions with technology.

3. Monitoring ongoing consent without any guidance.

As a senior researcher said, "You’re assessing if it’s still feasible rather than people having to indicate they can’t participate". So you have to observe whether these participants still want to be there. And then, as another senior researcher said, “You have to decide using your own judgment”. Do you stop the study, do you pause it, or do you continue? He continued to say that this relies on each individual researcher's moral compass, and this can be skewed through external pressures. Like when “a student who wants to graduate”, or “a professor who wants to apply for a fancy grant”.

4. Improving the consent process while adhering to the guidelines of the ethical review board (ERB).

All of the participants in this study recognised that we need individualised, personalised ways of working with people with dementia, and that the consent process we use for other people might not work for this population. Yet the senior researcher said that these forms “almost feels like a contract, you would not redesign a contract”. So his design students use the templates that the university provides, which are definitely not accessible to people with dementia. He also said that “that student said ‘we cannot get approval from the ERB board, so we will do research with experts’”. Because they perceive this as such a difficult process, they stop involving participants with dementia in their design research, and design technologies for them without actually involving them. Which is of course completely the opposite of what we are trying to do here.

A question for you

I want to reflect this back to you for a second. What do you as a researcher struggle with in your own research? This can be outside of the dementia context. Do you recognise that all of this study information can be overwhelming? Do you recognise that it can be frightening to give a signature or to document the agreement? Do you recognise that it is difficult to monitor ongoing consent and to make decisions based on it? And do you recognise how difficult it is to improve the consent process while adhering to the ERB guidelines?

The cause: consent as one-way communication

When we continued analysing this, we found that these issues arise when we treat consent as a one-way conversation. What do I mean by that? We see the study information as just going from the researcher to the participant. The signature just goes from the participant back to the researcher. The dissent signals just go from the participant to the researcher as well. And the ERB guidelines and templates are imposed on the researcher from above.

But if we look at communication theory, this is not the only way that communication works. Older models indeed thought that information just goes from the sender to the receiver. But newer ones consider communication more as a two-way street. I say something to you, you give me feedback on whether you understand it or agree with it, and then I adjust my message again, and that is how we keep communicating. Moreover, the context is hugely important in this conversation, meaning the physical, psychological, social, relational and cultural context. All of this can hinder or improve the conversation we are having.

Dynamic and Contextual communication

So we went back to the transcripts and looked at which strategies people were using. A lot of them were linear communication strategies, such as simplifying information or adding icons to the consent forms. But a lot of it was actually dynamic as well: checking whether the participant had understood the information, or involving them in the decision whether to stop or continue in the study. And some were using contextual communication as well. For example, creating a room with fewer distractions so it is easier to take in the study information, or involving another person who already has a relationship with the participant, so they can help communicate consent.

Looking back at my own interview

Let us look back at the moment that I described in the beginning.

I think I used linear communication. I used very simple language myself and handmade this simplified consent form.

I had also included dynamic communication, by checking with every consent item whether he had understood it, and asking him whether he wanted to continue participating or stop the study.

What I had not thought of was the context.

I had done this study in a separate room in the meeting center. I did this because I thought it would reduce distractions and create a clearer audio recording. What I had not considered was how this room would make someone feel. I had separated the person from the rest of the group, and he could not see or communicate with them. And I was sitting in front of the door.

So if this participant had wanted to leave, he would have had to get up, ask me to move, open the door, and then leave. That is a lot of steps for someone who finds it difficult to say no.

What I could have done better was not sit in front of that door, or leave the door open, or not do it in a separate room at all. Anything so that the participant with dementia could leave without having to explicitly state it in the interview.

How can we improve this?

How can we improve all of this within dementia research, and also outside of it?

1. Plan your consent conversations across all four challenges.

When we plan our consent conversations, we should think about the four different challenges that can occur, and think about which types of communication we can involve. We have made a set of reflective questions based on the strategies used by participants, and we put those in our paper.

2. Practise your communication skills.

As you saw, a lot is put on the researcher: to convey the information, to make the conversation a comfortable experience, and to check whether the other person still wants to be there. This is communication skills. We need to practise these four different challenges.

3. Make consent materials dynamic.

We are using static forms that try to convey information in one go. How about we use fold-out forms that adjust the level of detail to the participant's wish? Or, instead of the static signature, more of a handshake, something that reduces the fright. We have more ideas in the paper to make more dynamic consent materials. Let us be more creative with all of this.

4. Share what goes right and wrong with your fellow researchers.

As we said, context plays a huge part in these conversations, and we can only understand this context by sharing contextualised experiences. As I shared my experience in the beginning, only through sharing this can we learn from each other's mistakes. It feels uncomfortable to do this. It feels vulnerable to describe how something went wrong. But that is the only way that we can learn.

In short

Remember: consent is not just a signature, it is an entire conversation. To help you with your consent process, we added the following to our paper; the strategies used by our participants, a list of reflective questions to help you plan your consent process, and the consent form template that we used in our study.