John Gruber posted at Daring Fireball that when he washes his hands, he turns the water off and feels less rushed and more likely to wash for the CDC suggested twenty seconds. “It’s very clear to me after just two days that doing so makes it far more natural to spend more time actually sudsing your hands up. When you leave the water running, it subconsciously puts you in a bit of a rush, because you know you’re wasting water.”
Rationally whether the water runs or not shouldn’t matter. The most important thing (mid-March 2020) is to kill the harmful viruses and bacteria people pick up during their (limited) social exposures. Though the chances are small, the consequences are the largest. However we aren’t rational and we don’t always wash our hands.
At one teaching hospital, the best predictor of hand-washing was attending physicians. If they washed, the medical students followed. Multiple meta-analysis (meta-meta-analysis?) suggest the best option might be “multifaceted” nudges, educational materials, and bedside hand sanitizers. Another showed that performance reviews (personal wealth) and access to hand sanitizer (ease) had the strongest though-not-super-duper-strong effects. Incentives (personal health) also kept hand-washing levels high after the 2003 SARS outbreak.
What’s so interesting is that even though one path is clearly better, people need help following it. Hygiene is like diet or investments.
This randomized control trial in India found a way to increase hand-washing 30X, even twelve months after the intervention.
A study of 802 Kenyan households offers the model that makes the most sense to me for why people do anything. Those, “significant predictors of observed hand-washing behaviour: having the habit of hand-washing at particular junctures during the day, the motivated need for personal or household cleanliness, and a lack of cognitive concern about the cost of soap use.”
Like finches, people are influenced by their environment. If we want to encourage actions like hand washing, social distancing, and factfullness we should design conditions that make those thing easy.