Picture a health leaflet with a maze: boxes for sleep, stress, food and exercise, with arrows running between them. That's a lot of information! Extra detail may be accurate, but does not mean it helps you decide what to do next.
In an online experiment with 1,077 U.S. adults, small diagrams helped people choose answers to hypothetical questions about managing Type 2 diabetes. As the diagrams grew, that advantage shrank. The largest diagrams did not produce significantly better scores than showing no diagram at all.
Greta Mohr and Samantha Kleinberg randomly assigned participants to five groups. One received no diagram; the others saw diagrams containing two, four, eight or 16 labeled elements connected by arrows. Each person answered five multiple-choice questions concerning physical activity, stress, cholesterol, blood pressure and sleep.
One fictional scenario described a man who worked from home, rarely exercised and relied on fast food. He wanted to manage his diabetes without joining a gym or cooking at home. The researchers designated short daily walks as the best of four supplied options.
The smallest accompanying diagram connected lack of physical activity directly to healthy blood sugar. Larger versions retained that relevant connection while adding other factors and arrows. Crucially, the two-element diagrams contained only the path relevant to the correct answer. The experiment therefore tested the cost of adding information around an already useful cue.
People shown two- or four-element diagrams outperformed those shown eight or 16 elements, as well as those given no diagram. The researchers found no evidence that greater factual knowledge protected people from the cost of added complexity.
The study says that it did not establish that every complex health decision can safely be reduced to two boxes, but that added information has a cost, not always a benefit.
“prioritise clarity and focus over comprehensiveness”— The authors write in Discussion, “The benefit and cost of causal diagrams,” fourth paragraph; quotation from the publication
The research
Greta Mohr and Samantha Kleinberg
Cognitive Research: Principles and Implications · Accepted
DOI: 10.1186/s41235-026-00756-4