Same Choices, Different Words: Why Framing Matters in Group Decisions

A team is considering a new customer support process. According to the estimates, it would resolve 80 out of every 100 cases.

One person describes that as 80 customers getting the help they need. Another points out that 20 would still be left without a solution.

They're working with the same numbers. Yet one description draws attention to the people who benefit, while the other emphasizes those who don't.

In everyday discussions, people may also disagree about the facts, the risks, or what matters most. But even when the outcomes are identical, the way they're described can influence which option people choose.

Two Ways to Describe the Same Decision

In 1981, psychologists Amos Tversky and Daniel Kahneman asked university students to imagine an unusual disease outbreak expected to kill 600 people. Two programs had been proposed to combat it, and the students had to choose one.

Program A would save 200 people for certain.

Program B offered a one-in-three chance of saving all 600 people, but a two-in-three chance that nobody would be saved.

Both programs offered the same expected number of survivors. The difference was certainty. One guaranteed that 200 people would live, while the other offered a chance to save everyone, with the risk of saving no one.

Of the 152 students who received this version, 72% chose Program A. Most preferred the guaranteed outcome.

Another 155 students received a different version of the problem.

This time, Program C would result in 400 people dying.

Program D offered a one-in-three chance that nobody would die, but a two-in-three chance that all 600 would die.

Now 78% chose Program D, preferring to take the risk rather than accept the certain death of 400 people.

Yet the options were exactly the same.

Saving 200 people also means that 400 will die. A one-in-three chance of saving everyone is the same as a one-in-three chance that nobody will die.

Nothing about the possible outcomes or their probabilities had changed. Only their descriptions were different.

When the programs were presented in terms of lives saved, most participants preferred certainty. When presented in terms of lives lost, most preferred taking a risk.

This is an example of the framing effect. The way equivalent options are described can influence which one people prefer.

The researchers had given the two versions to different participants. They hadn't shown that the same person would necessarily reverse their decision. But the difference between the two groups was striking: describing identical outcomes as gains or losses produced very different patterns of choice. ¹

Would a Group Make the Same Choice?

The students in the original experiment made their decisions individually. They couldn't discuss the problem with someone who might question the wording or suggest another interpretation.

A group might bring different perspectives to the table. Someone could point out that saving 200 people also means losing 400. Another might ask whether the risky program is really more attractive simply because the certain outcome sounds unpleasant.

But having people discuss a decision doesn't guarantee that they'll reconsider how it was presented.

In 2009, Kerry Milch and colleagues studied how groups responded to framing. They recruited 102 people who already knew one another through university organizations and workplace teams, forming 34 groups of three.

One of their tasks was a modified version of the original disease problem. This time, an outbreak of West Nile virus threatened 600 students at Columbia University. Participants had to choose between two proposed vaccine research programs.

As before, some received descriptions emphasizing how many people would be protected, while others saw how many would become infected.

Every member of a given group received the same wording. Some participants considered the decision individually before their group discussed it. Others encountered that particular problem for the first time during the group discussion.

Among those making an initial individual decision, 80% chose the risky option when the outcomes were described as losses, compared with 50% when they were described as gains.

The group decisions followed a similar pattern: 82% chose the risky option in the loss condition, compared with 56% in the gain condition.

The difference between the two framing conditions among groups was not statistically significant, and the researchers did not find clear evidence that group discussion reduced framing effects compared with individual decisions.

This doesn't establish that groups are equally influenced by framing. With only 34 groups, the findings leave considerable uncertainty. But they provide no reliable support for the assumption that discussing a choice together automatically removes the influence of its wording.

The researchers also examined a choice between receiving a smaller amount of money sooner or a larger amount later. The results were different. Groups encountering the choice together for the first time showed a framing pattern opposite to one reported in earlier individual research, while groups whose members had already considered the decision showed no significant framing effect.

The financial task differed from the disease problem, including in how its alternatives were constructed. Its results reinforced the need for caution when drawing broad conclusions about how groups respond to framing. ²

The Reasons Behind the Choice

Milch and colleagues didn't just record which programs the groups selected. They also collected written explanations and recorded the conversations leading up to each decision.

This allowed them to examine what participants actually discussed.

In the disease problem, one recurring consideration was certainty. Some participants emphasized the value of a guaranteed outcome, knowing that a particular number of people would be protected rather than taking a chance on saving everyone.

References to certainty were associated with choosing the guaranteed option.

Other arguments focused on broader social goals or compared the alternatives analytically. In the group discussions, these kinds of arguments were associated with choosing the riskier program.

The researchers found that the reasons participants gave helped explain their choices beyond knowing which framing condition they had received.

That doesn't mean particular arguments caused the decisions. Participants may have been explaining preferences they already held, and the researchers did not establish a consistent causal mechanism connecting the framing, the arguments, and the final choices.

Still, the conversations added something important.

Knowing which program a group selected tells us the outcome of its decision. Hearing the discussion gives us a clearer picture of what people considered important when they made it.

A group choosing the guaranteed program might have valued certainty above the possibility of saving everyone. A group selecting the riskier option might have been more concerned about accepting a limited outcome when a better one was still possible.

And even when people reach the same decision, they may have arrived there for different reasons.

For a facilitator, those reasons are worth exploring before the group moves on.

Before the Group Moves On

Imagine a workshop where a team has to prioritize possible responses to a project that has fallen behind schedule.

After debating several options, the participants agree to place reducing the project's scope near the top of their list.

Before moving on, the facilitator asks:

"What mattered most to us when we made that choice?"

One participant says reducing scope will help the team meet its deadline. Another believes it will protect the quality of the remaining work. A third is concerned about the pressure on an already exhausted team.

They've agreed on a priority, but their reasons aren't quite the same.

The facilitator then invites the group to look at the decision differently.

Suppose reducing scope means the next release will contain eight of the ten originally planned features. That outcome can be described in two ways:

"We'll still deliver eight of the ten planned features."

Or:

"We'll have to leave out two of the ten planned features."

The facts haven't changed. One description emphasizes what the team will deliver. The other draws attention to what customers won't receive.

Does either description make the decision feel more or less acceptable? Does it bring a consequence into focus that the group hadn't discussed?

The team may still believe reducing scope is the right choice. But comparing the two descriptions gives participants a reason to examine what they were focusing on when they agreed.

This is a practical suggestion inspired by framing research, not a technique validated by these experiments. The researchers presented equivalent descriptions to different participants. They didn't test whether openly comparing those descriptions in a facilitated discussion would improve a group's decision.

And in a real workshop, participants may genuinely disagree about the likely outcomes or which features matter most. Those differences aren't necessarily framing effects. They are additional reasons to examine the thinking behind a decision.

In a facilitated activity such as Protocol15, teams negotiate a shared ranking of possible actions or priorities. The completed ranking and its comparison with a reference guide offer a concrete starting point for reflection. They show which options the group prioritized, but not why those choices felt right.

The group has made its decision. The facilitator can still ask what, exactly, everyone believes they've agreed to.