6 ideas
Behaviour is often blocked not by weak intentions but by the absence of a concrete route from intention to action. In Leventhal's tetanus study, frightening health messages changed attitudes but not behaviour. Giving students a campus map with the clinic circled and asking them to pick a time raised vaccination from about 3% to 28%.
In Darley and Batson's Good Samaritan study, seminary students walked past a groaning man on their way to give a talk. Whether they stopped depended mainly on how late they had been told they were: about 63% of unhurried students helped, versus about 10% of hurried ones. Their measured religiosity predicted almost nothing, and some who passed him were on their way to speak on the Good Samaritan parable.
Correlations between a measured trait and behaviour in any specific new situation rarely exceed about .30. Correlations between the same trait-relevant behaviour across two different situations often sit near .10–.20. As a result, knowing someone's past honesty or friendliness in one setting tells you surprisingly little about how they will act in the next.
Observers usually see others in a narrow band of recurring roles and settings, often ones that include the observer. Genuine predictability tends to come from stable construals and chosen environments rather than from broad traits like 'conscientious' or 'aggressive'.
People respond to their subjective interpretation of a situation, not to its objective features. When behaviour seems bizarre or trait-driven, it often makes sense once you learn how the actor was defining the situation. Misunderstandings persist because each party assumes their own construal is simply how things are.
Individuals and groups are held in equilibrium by opposing forces, such as group norms, identities and institutional pressures. A modest intervention can fail entirely if it pushes against a strong restraining force. A tiny change can cascade if it releases a force that was holding a stable state in place, so the effect of an intervention depends on the surrounding system of forces rather than on its size.