EV-0048
If-then plans, and why every practice should be written as one
Gollwitzer, P. M., & Sheeran, P. (2006). Advances in Experimental Social Psychology, 38, 69-119 · Source
What this entitles us to say
A meta-analysis of 94 independent tests finding a medium-to-large effect (d = .65) of implementation intentions - specific if-then plans - on goal attainment, operating by initiating action, shielding it from interference, enabling disengagement from failing courses, and conserving capacity.
What it found
94 independent tests. Implementation intentions produced a medium-to-large effect on goal attainment, d = .65. Four mechanisms were identified: they promote initiation of goal striving, shield ongoing pursuit from unwanted influences, enable disengagement from failing courses of action, and conserve capability for future striving.
Why it matters here
This is the practical core of A16 and it should govern how the practice library is written.
The form is the finding. A practice phrased as an aspiration - be more patient, listen better, keep your word - is an intention. A practice phrased as "when [specific situation], I will [specific action]" is an implementation intention, and it is the version with d = .65 behind it. Every entry in the practice library should be written in the second form, and that is a rule the research supports rather than a stylistic preference.
Two of the four mechanisms connect to traits already in the library. Shielding is what makes Discipline work through arrangement rather than willpower (EV-0027). Disengagement from failing courses of action is the antidote to escalation of commitment (EV-0025) - a leader who has pre-committed to a stopping condition is far better placed than one relying on judgment at the moment their ego is on the line.
That last point is worth making explicit in the Commitment module: the time to decide when you will stop is before you start, and there is a mechanism-level reason it works.
The limits
The underlying studies are largely on health, academic and personal goals rather than on character or leadership behavior - the transfer is by analogy, and a good one, but ours. Effects are strongest for goals people already hold; if-then planning does not create the wish to change, it converts an existing one into action. Publication bias is a live concern in this literature as in any with a large positive headline effect.
What we do not say
that this was demonstrated on character or leadership behavior. The underlying studies are largely health, academic and personal goals. The application here is ours.