Mechanisms — implementation intentions IN DEPTH (process model, domains, limits)

Overview

An in-depth analysis of implementation intentions ("if situation X — then action Y"; Gollwitzer) beyond the basic meta-analysis already recorded in intention as a program (there the summary effect is tied to Wang, Wang & Gai 2021, a meta-analysis of MCII). Three additional questions, each tied where possible to a genuinely open-access peer-reviewed source:

1. MECHANISM — why "if X — then Y" works. Gollwitzer's process model: forming the plan raises the accessibility / perceptual readiness for the cue and builds a strong associative link cue→action, so that the action is triggered by "strategic automaticity" — efficiently, immediately, without fresh conscious deliberation. The practitioner delegates control of the action in advance to a pre-chosen situational cue (bottom-up rather than top-down control).
2. BREADTH AND STRENGTH across domains — where the effect is larger and where smaller (health, study, prosocial/environmental behavior, clinical), with effect sizes, from open meta-analyses where they exist and from (non-open) named meta-analyses cited second-hand and flagged.
3. LIMITS — when it does NOT work — weak/absent goal commitment, a vague or indistinct cue, competing entrenched habits, overuse / a prospective-memory cost, ceiling effects.

The strength of evidence is stated per source. Lab / meta-analytic / single-domain coverage → mechanism plausibility, not proof of "manifestation" at the population level. Generalization to "manifestation" / life events stays [unverified].

1. The process model: cue accessibility + strategic automaticity (MECHANISM)

Citation (main open source): Wieber, F., Thürmer, J. L., & Gollwitzer, P. M. (2015). Promoting the translation of intentions into action by implementation intentions: behavioral effects and physiological correlates. Frontiers in Human Neuroscience, 9, 395.

• mechanism-evidence Raised cue accessibility / perceptual readiness. Forming an implementation intention "makes the mental representation of the situation highly accessible" — because the practitioner pre-selects a concrete future situation, its representation is activated and more easily noticed. Priming studies (Aarts et al. 1999, cited in the article) show faster responses to the critical cue among those who form if-then plans. → The "if X" half preloads perception so the cue is caught when it appears. This is exactly the function of the chosen cue of a runescript (the moment, object, or sensation the practitioner ties the stave to).
• mechanism-evidence A strong associative cue→action link. The plan also builds "a strong associative link between the mental representations of the situation and the action." The cue and the response are linked, so meeting the cue tends to elicit the response. → The "then Y" half is not a fresh decision each time; it is pre-soldered to the cue.
• mechanism-evidence Strategic automaticity = delegating control to the cue. The article describes the effect as achieving "strategic automaticity of action control… by switching from top-down to bottom-up information processing": the person deliberately delegates control of goal-directed behavior to pre-chosen situational cues. The noted features of this automaticity: efficiency (responses survive a dual load), immediacy (fast triggering), and redundancy of conscious intention (even subliminal / below-awareness cues can trigger the response). → Mechanistically, this is the strongest formulation of "the program runs itself": the action is triggered by the situation, not by renewed willpower. Note: "automatic" here = cue-triggered, and NOT "without prior intention" — the plan was deliberately formed first.
• mechanism-evidence Physiological / neural correlates (not magic, biology).
  • fMRI: acting on implementation intentions engaged the medial rostral prefrontal cortex (medial BA 10) — linked to stimulus-driven control — whereas plain goal intentions engaged the lateral BA 10 / dorsolateral PFC (effortful control). Emotion regulation via if-then plans engaged the OFC with reduced amygdala activity.
  • EEG: spider-fear regulation showed a smaller occipital P100 (early visual modulation); a go/no-go study in children with ADHD showed P300 amplitude differences (doubled), consistent with strategic automaticity.
  • Stress physiology: under acute stress (raised cortisol + heart rate, Trier Social Stress Test) performance on an implementation intention was preserved where performance on a plain goal intention was not. → If-then plans appear robust to stress load — relevant to the claim "the stave works even when I'm anxious."
• mechanism-evidence The article itself notes the limits of "automaticity." Implementation intentions also slowed the ongoing task (a prospective-memory cost, d ≈ 0.22) — so the process "is not purely automatic but involves a complex interplay of automatic and controlled processes." (Carried into §3.)

2. Breadth and strength across domains (BREADTH AND STRENGTH)

Citation (main open source — the domain pillar): Silva, M. A. V., São-João, T. M., Brizon, V. C., Franco, D. H., & Mialhe, F. L. (2018). Impact of implementation intentions on physical activity practice in adults: A systematic review and meta-analysis of randomized clinical trials. PLoS ONE, 13(11), e0206294.

• mechanism-evidence Health (physical activity) — modest, conditional. 13 RCTs (11 in the meta-analysis), sample sizes 30–709. Overall SMD = 0.15, 95% CI (−0.01, 0.31) — NOT significant; heterogeneity I² ≈ 54% (moderate). The effect became significant only with reinforcement of the plans during follow-up (by phone/SMS/in person): SMD = 0.25, 95% CI (0.05, 0.45); without reinforcement — not significant. Studies without coping planning (strategies for obstacles) showed non-significant differences. → For unsupported, repeated health behavior, a bare if-then plan is weak; it needs reinforcement and coping plans. (Flows straight into the limits of §3.)

The domain breakdown — given via the non-open Gollwitzer & Sheeran 2006. ⚠️ The widely cited per-domain effect sizes originate in Gollwitzer & Sheeran (2006) (an Elsevier book series, paywalled) — the figures below are given as a paraphrase via the secondary literature:

• mechanism-evidence (via Gollwitzer & Sheeran 2006, not open) Across 94 independent tests (~8,000+ participants) implementation intentions gave a medium-to-large overall effect, d ≈ 0.65. The often-cited domain picture: large for anti-racist / prosocial / environmental behavior (d ≈ 0.87 / 1.01 / 1.12), medium-to-large for lab tasks and academic performance (d ≈ 0.70 / 0.72), medium for consumer behavior, health behavior, and personal goals. → A reading of the picture: larger where the behavior is discrete, novel, and triggered once (a single anti-racist response, a single recycling act, sitting down to study); smaller where the behavior is repeated, habitual, or limited by skill/endurance (sustained workouts) — consistent with the open Silva 2018 finding above.
• mechanism-evidence (via Toli, Webb & Hardy 2016, British Journal of Clinical Psychology, 55, 69–90 — paywalled) Across 28 experimental studies forming implementation intentions gave a large effect, d+ ≈ 0.99 on goal attainment in people with diverse mental-health difficulties. → It suggests the technique can be strong in clinical/analog goal attainment too, but the number is from a closed source. Toli → negative.

Strength of evidence: The open-access-defensible domain number here is conservative, conditional: Silva 2018 (CC BY): bare if-then plans for sustained physical activity are weak/non-significant unless reinforced. The large, flattering domain figures (d ≈ 0.65 overall; 0.7–1.1 for study / prosocial / clinical) belong to closed meta-analyses (Gollwitzer & Sheeran 2006; Toli et al. 2016) and are given here only second-hand. Bottom line for the knowledge base's honesty: the effect is real and genuinely broad across domains, but the largest numbers sit in closed sources, while the independent open evidence in the hard (repeated behavior) domain is modest and conditional. The honest headline number: stronger for discrete/novel actions (a study session, a single decision), weaker for goals requiring sustained habits (regular workouts) unless reinforced. [unverified] transfer of any of these domain effects to "a runescript for a life outcome."

3. Limits: when it does NOT work (LIMITS)

This section rests on the open Silva 2018 (CC BY) and open Wieber 2015 (CC BY) findings above, plus a named but not open goal-commitment demonstration, honestly flagged.

• mechanism-evidence Weak / absent / unactivated goal commitment — the decisive limit. (via Sheeran, Webb & Gollwitzer 2005, The interplay between goal intentions and implementation intentions, PSPB, 31(1), 87–98 — paywalled). Their experiments showed implementation intentions improved goal attainment only when the underlying goal intention was strong, and (Study 2) only when the relevant goal was activated (even unconsciously) — no benefit with weak goal commitment or an inactive goal. → For a runescript: an if-then stave tied to a goal the practitioner is not actually committed to should do little. The plan is an execution aid; it does not produce motivation. Sheeran et al. 2005 → negative (cited bibliographically).
• mechanism-evidence A vague / indistinct / irrelevant cue. From Silva 2018 (CC BY) + the general literature: bare plans for frequent, repeated behavior (e.g. everyday workouts) underperform, because with many opportunities to act the situational cue loses distinctiveness and stops triggering the action. A cue that is ambiguous, rarely met, or not actually tied to the action breaks the cue→action link the mechanism rests on. → A runescript cue that is blurry or omnipresent ("whenever I think of it") is weaker than a concrete, salient, reliably occurring cue.
• mechanism-evidence Competing / entrenched habits and skill/endurance limits. Implementation intentions work best when the barrier is getting started (a clear initiation problem), and worst when the barrier is skill, endurance, or low commitment. A strong existing habit can override the planned cue→action link (the old automatic response fires first). → If the obstacle is "I don't know how" or "I can't sustain it," an if-then plan alone won't fix it; this is where coping planning (Silva 2018: its absence → non-significant effects) and reinforcement matter.
• mechanism-evidence Overuse / cognitive cost (the technique isn't free). Wieber 2015 (CC BY) reports that holding an implementation intention slowed the ongoing task (a prospective-memory cost, d ≈ 0.22) and that the process "is not purely automatic." → Stacking many simultaneous if-then plans (or many runescripts) plausibly blurs each and adds monitoring load — more is not strictly better. direct evidence on overuse of the number of plans was not extracted.
• mechanism-evidence Ceiling effects. Wieber 2015 (CC BY) notes a study where older adults who were already doing well got no additional benefit from implementation intentions. → When behavior is already near optimal, the plan has little to add.

Strength of evidence: The goal-commitment limit — the single most important "when it fails" condition — is demonstrated in a closed source (Sheeran et al. 2005, PSPB), cited here bibliographically; the limits on cue distinctiveness / coping planning / reinforcement and the limits on prospective-memory cost / ceiling are backed by the open Silva 2018 and Wieber 2015. Bottom line for the knowledge base: implementation intentions are not a universal switch; they predictably fail under weak commitment, blurry/over-habitual cues, competing habits, skill/endurance barriers, and overuse — and these failure modes are honest, internal/behavioral facts, not proof for or against any "energy" claim.

Connection to the runescript

This file extends, not replaces intention as a program — that file established the summary effect (MCII g≈0.34 bias-corrected, via CC-BY Wang 2021; the classic d≈0.65 via non-OA second-hand) and the broader decomposition structured intention → if-then plan + a concrete goal + expectation → action → outcome. This file goes inside the term "if-then plan": the process, the domain spread, and the failure modes.

The runescript as a fixed cue→action pre-wiring. On this evidence, a stave "works" to the extent that it = (a) a chosen, distinct cue (if X) + (b) an action pre-wired to it (then Y) + (c) real goal commitment. The mechanism is internal/behavioral (raised cue readiness + strategic automaticity + delegating the trigger to the situation), not external "magic."

[unverified] any arrows of "→ a change in external physical reality without the practitioner's mediating action." "Strategic automaticity" means the action is triggered by the situation, NOT "a wish fulfilled without action." Transfer of the process model to "manifestation" (an outcome without cue-triggered behavior) is [unverified].