Why Good Intentions Often Fail

The gap between intending and doing

Most of us have resolved to do something and then not done it. That is an ordinary experience, not a personal failing. Psychologists have tested this gap directly. Webb and Sheeran reviewed experiments in which participants were randomly assigned to a treatment that significantly strengthened their intentions, and whose later behaviour was compared with a control group. They found 47 such tests.

Across those tests, a medium to large change in intention, measured at d = 0.66, led to a small to medium change in behaviour, measured at d = 0.36. The letter d is a standard way of expressing the size of an effect. We read this as follows: stronger intentions helped, but by less than many people assume.

What a habit study adds

Lally and colleagues asked 96 volunteers to choose an eating, drinking or activity behaviour and perform it daily, in the same context, for 12 weeks. Of these, 82 provided enough data for analysis. The time taken to reach 95 per cent of each person's asymptote of automaticity ranged from 18 to 254 days. The upper figure is a model estimate beyond the period actually observed.

The authors also report that missing one opportunity to perform the behaviour did not materially affect the habit formation process. A single lapse, in other words, was not the end of the effort.

What we take from this

We think the lesson is modest. Intention is a beginning, not an outcome. A useful question for any reader is whether a change has been made easy to repeat in the same setting each day. The habit study asked volunteers to act in the same context, which fits that idea, though it did not test it. Plans that rely on one burst of resolve are, in our view, optimistic. The range in the habit study is wide, and it suggests that the time a change takes differs a great deal from one person to the next.

There is also much these studies cannot tell us. Volunteers chose their own behaviours and the sample was small, so we treat the figures as indicative, not universal. Neither study tells any one reader which method will work for them.

We would rather report a modest finding accurately than a dramatic one loosely. We link the studies so that readers can open them. If we have read either one wrongly, we would like to be told.

Informational only, not medical advice.