A small body of research has looked at what happens when adults sit down and talk about death in a group, sometimes with people they’ve never met. In the studies below, players rated these talks as satisfying, their mood didn’t drop in a way the researchers could measure, and the talk stayed on people, values, and stories. The studies are small, and none of them compared players with a group who didn’t play.
This page is for hosts of death cafés, grief groups, and conversation evenings. A guest asks whether the evening will leave them feeling worse, or a venue asks why anyone would come. Below are the numbers you can quote, where they come from, and what they don’t prove.
How often people talk about death at all
Two surveys give a baseline for the US and the UK. In a January 2018 survey of 1,004 US adults, 32% had talked with loved ones about their wishes for end-of-life care, and 68% had not. A primary reason given was that it felt “too soon.” In the same survey, 53% said they’d feel relieved if a loved one started the conversation (The Conversation Project 2018).
In Britain, a ComRes survey for the Dying Matters Coalition, published in May 2012, found that 71% of the public agreed that people in Britain are uncomfortable discussing dying, death, and bereavement. In the same survey, 31% had talked about their own end-of-life wishes, and 27% had asked a family member about theirs (Dying Matters / ComRes 2012). The UK numbers are over ten years old, so quote them with the year.
Together, the two surveys describe a gap. Many adults see discomfort around the topic, fewer than a third have raised their own wishes, and about half of the US sample would welcome someone else going first.
What players said after a conversation game
The most direct evidence comes from a research team that tested a conversation game about death. In the first study, 70 healthy volunteers played in 18 groups made up of families, friends, or strangers. Each group discussed 20 questions about death, dying, and end-of-life care over about two hours. On a seven-point scale, players rated their conversation satisfaction at a mean of 6.1, realism at 5.6, and the quality of their own talk at 5.7. Their emotional state after the game wasn’t noticeably worse. In focus groups, players called the game enjoyable and useful, and they had mixed opinions about who the ideal group would be (Van Scoy et al. 2016a).
Those mixed opinions matter for hosts. The study included groups of strangers, and it didn’t settle whether strangers, friends, or family make the better table. A room of strangers was one of the tested formats.
What the talk was about
A second paper from the same team recorded 17 games with 2–6 players each, 68 people in total, played without a facilitator. The researchers transcribed the games and sorted what people said. Four themes came up: the role of people and relationships in end-of-life decisions, values and preferences about care, preparing for what happens after a death, and using stories and past experiences to shape one’s own wishes (Van Scoy et al. 2016b).
For a host, this answers a quieter worry: that a table left alone will drift into small talk or into medical detail nobody can answer. In these recordings, people talked about each other and about what they’d lived through. Stories and past experiences were one of the four themes.
What people did in the months after
A 2020 study followed African American adults who played the same kind of game in groups of 4–6 for 60 minutes. Of 220 people reached at follow-up, 3–11 months later, 80% (176) had discussed their end-of-life wishes with loved ones, and 41% (91) had completed a new advance directive. The mean conversation satisfaction score was 6.21 out of 7, and no participants withdrew because of an adverse event (Van Scoy et al. 2020).
The design was a cohort study. It followed one group of players over time and didn’t randomize anyone, so it can’t show that the game caused these conversations. People who sign up for an evening about death may already be ready to have them. That’s the honest limit to state when you quote the 80%.
Why strangers can be easier than expected
A separate line of research looks at talking with strangers in general. In a series of experiments, Kardas, Kumar, and Epley had people hold deep or shallow conversations. Conversations with strangers felt less awkward and left people more connected and happier than they had expected. Participants overestimated how awkward deep conversations would be, and they felt more connected to deep conversation partners than to shallow ones. The gap came from one belief: people expected strangers to care less about their disclosures than the strangers did (Kardas et al. 2021).
This study wasn’t about death. It does describe the guest who almost didn’t come in, sure that nobody wants to hear about their father’s last year. The research suggests the people across the table may be more interested than that guest expects.
What the research can’t tell you
The evidence has clear edges. Five limits apply to the numbers above:
- Sample sizes. The game studies had 70, 68, and 380 participants.
- Self-selection. Volunteers who agree to talk about death may differ from guests who stay home.
- One game, one team. The three game studies come from the same research group and similar game formats.
- No comparison group. None of the game studies compared players with people who didn’t play.
- Survey years. The US survey is from 2018 and the UK survey from 2012.
None of this measures grief, mental health, or recovery, and a conversation evening is not a clinical service. If someone at your table needs more support than a group can give, point them to a professional.
Three words that fit a table of strangers
The studies above used question cards. A single word can do the same job with less steering, because each guest decides where it goes. Thesis cards have an idea on one side and an everyday thing on the other.
Death has Phone on its other side. A phone holds the call that brought the news, a voicemail someone kept, and a contact nobody deleted. A guest who isn’t ready for the idea side can start with the phone, and the story can reach the same place.
Regret has Lamp on its other side. Regret ties into the “stories and past experiences” theme from the 2016 content analysis. The lamp is gentler: whose desk it stood on, who left it on at night.
Home has Spoon on its other side. Where people want to be at the end is a question about home. A spoon brings up a kitchen, a recipe, and who did the feeding, which is an easier first answer for someone who arrived a few minutes ago.
Trying it at your next session
Before the round starts, read out the line from the Thesis rules: “Before you start, agree together how passing works.” Passing lets a newcomer listen for a round before saying anything. Then draw a card, choose a side, and roll the die. Its faces are MEMORIES · BODY · PEOPLE · VALUES · CHILDHOOD · FUTURE. “Look at the word from that side.”
After an answer, the rule card offers three follow-ups:
What did that feel like? Who else was there? Has it always meant that to you?
For prompts at a Death Cafe, see death cafe conversation starters. For more on groups that talk about loss, see Thesis for therapy groups.
At your next session, take one number from this page to the guest who asks whether the evening will leave them feeling worse: 70 volunteers rated their conversations in a game like this at 6.1 out of 7 for satisfaction. Then tell them they can pass.
The word pages for Death, Regret, and Home have more on each card. The research behind the deck explains where the words came from. The online table is free to use, and the rules and the box are on the product page. For the game studies in more detail, see the end of life conversation game studies.