The effectiveness of implementation intentions in children: A systematic review and meta‐analysis

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

As the frontal lobe and associated cognitive abilities continue to develop into late adolescence, children have a harder time focusing their attention on goal-relevant cues, shifting flexibly between different goals and inhibiting competing impulses.

“As the frontal lobe and associated cognitive abilities continue to develop into late adolescence (Fuster, 2002), children have a harder time focusing their attention on goal‐relevant cues, shifting flexibly between different goals and inhibiting competing impulses (Diamond, 2013; Fuster, 2002; Luna et al., 2004).”

• cites Diamond, 2013 sourceSupported
• cites Fuster, 2002 sourceNot checked — no full text of the source
• cites Luna et al., 2004 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Children who are better at self-regulation in early childhood are more likely to excel academically and socially, leading healthier lives with fewer behavioural problems.

“Children who are better at self‐regulation in early childhood are more likely to excel academically and socially, leading healthier lives with fewer behavioural problems (Duckworth et al., 2019; Moffitt et al., 2011; Perry et al., 2018).”

• cites Duckworth et al., 2019 sourceNot checked — no full text of the source
• cites Moffitt et al., 2011 sourcePartly supported

  The Results report that greater childhood self-control predicted better health and fewer criminal outcomes in adulthood. The sibling comparison also links self-control at age 5 to better school performance and fewer antisocial behaviours at age 12. The paper does not establish that children excel socially or that academic and social success lead to healthier lives; these findings are predictive associations, not a demonstrated causal sequence. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites Perry et al., 2018Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Due to this link, the action is executed immediately, efficiently, independent of conscious intent and with reduced controllability.

“Due to this link, the action is executed immediately, efficiently, independent of conscious intent and with reduced controllability (Bayer et al., 2009; Brandstätter et al., 2001; Gollwitzer & Brandstätter, 1997; Wieber & Sassenberg, 2006).”

• cites Bayer et al., 2009 sourceNot checked — no full text of the source
• cites Brandstätter et al., 2001 sourceNot checked — no full text of the source
• cites Gollwitzer & Brandstätter, 1997 sourcePartly supported

  In “How Do Implementation Intentions Promote Goal Pursuit?” (page unavailable), the paper says the situation–behavior link can initiate action swiftly, effortlessly, and without a conscious intent; Study 3 reports more immediate action initiation. It does not establish reduced controllability. The “Similarities and Differences of Related Approaches Planning” section leaves open whether implementation intentions reduce flexibility. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Risk-taking behaviour increases in adolescence due to faster development of the socio-emotional system compared to cognitive control.

“At the same time, risk‐taking behaviour increases due to faster development of the socio‐emotional system compared to cognitive control (Steinberg, 2008).”

• cites Steinberg, 2008 sourceSupported

May combine two claims.

The importance of self-regulation for long-term development and its variability in children make self-regulation a main target for psychological interventions.

“Taken together, the importance of self‐regulation for long‐term development and its variability in children (Montroy et al., 2016) make self‐regulation a main target for psychological interventions.”

• cites Montroy et al., 2016 sourceSupported

Classified as probably not a substantial claim.

The authors independently assessed risk of bias using Cochrane Risk of Bias Tools (RoB 2.0, RoB 2 crossover, ROBINS-I).

“The authors independently assessed the risk of bias in the included studies using the Cochrane Risk of Bias Tools, which were suitable for each study (RoB 2.0, RoB 2 crossover, ROBINS‐I; Eldridge et al., 2016; Sterne et al., 2016, 2019).”

• cites Eldridge et al., 2016Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Sterne et al., 2016 sourceSupported
• cites Sterne et al., 2019 sourceSupported

May combine two claims.

Gawrilow and Gollwitzer (2008) found that only children with ADHD benefited from the intervention, while those without ADHD did not.

“Only children with ADHD benefited from the intervention, while those without ADHD did not.”

• cites Gawrilow and Gollwitzer (2008) sourceSupported

Classified as probably not a substantial claim. May combine two claims.

Implementation intentions are a simple strategy that helps people translate their good intentions into action.

“Implementation intentions have been introduced as a ‘simple’ strategy that helps people translate their good intentions into action (Gollwitzer, 1999).”

• cites Gollwitzer, 1999 sourceSupported

May combine two claims.

Effects have been demonstrated in patients with frontal lobe lesions and children with ADHD, suggesting that implementation intentions may be effective even in populations with suboptimal self-regulation abilities.

“On the one hand, effects have been demonstrated in patients with frontal lobe lesions (e.g. Lengfelder & Gollwitzer, 2001) and children with ADHD (e.g. Gawrilow & Gollwitzer, 2008), suggesting that implementation intentions may be effective even in populations with suboptimal self‐regulation abilities.”

• cites Lengfelder & Gollwitzer, 2001 sourceNot checked — no full text of the source
• cites Gawrilow & Gollwitzer, 2008 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

The absence of a significant decrease in effect sizes for studies with longer time intervals could suggest that implementation intentions may have lasting effects.

“The absence of a significant decrease in effect sizes for studies with longer time intervals could suggest that implementation intentions may have lasting effects, which is consistent with the idea that they promote instant habit formation (Gollwitzer, 1999).”

• cites Gollwitzer, 1999 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

The lack of variation in cue type likely reflects a shared understanding in the literature that cues should be concrete and salient.

“This lack of variation likely reflects a shared understanding in the literature—grounded in theoretical work by Gollwitzer (1999)—that cues should be concrete and salient (see also Schaaf et al., 2025).”

• cites Gollwitzer, 1999 sourcePartly supported

  In “The Specified Situation” (page unavailable), Gollwitzer describes plans that specify when and where to act and says specified cues become easier to detect and attract attention. This provides a theoretical basis for identifiable, noticeable cues, but the paper does not say that cues should be intrinsically concrete and salient. The proposed explanation for the lack of variation is the citing authors’ inference, not a finding attributed to Gollwitzer. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites Schaaf et al., 2025 sourceNot checked — no full text of the source

May combine two claims.

In studies with the longest follow-up periods, the intervention was typically reinforced through repeated sessions, which likely contributed to the sustained effects.

“Second, in studies with the longest follow‐up periods, the intervention was typically reinforced through repeated sessions (e.g. Conner et al., 2019; Conner & Higgins, 2010), which likely contributed to the sustained effects.”

• cites Conner et al., 2019 sourceSupported
• cites Conner & Higgins, 2010 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

Implementation intentions are only effective when the individual is committed to the underlying goal.

“Theoretical and empirical accounts emphasize that implementation intentions are only effective when the individual is committed to the underlying goal (Sheeran et al., 2005).”

• cites Sheeran et al., 2005 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

There is a long history in medicine of written action plans, mainly used in the treatment of asthma bronchiale, but also in the treatment of atopic dermatitis.

“For instance, there is a long history in medicine of written action plans, which have mainly been used in the treatment of asthma bronchiale (Zemek et al., 2008), but also in the treatment of atopic dermatitis (Thandi et al., 2023).”

• cites Zemek et al., 2008 sourceNot checked — no full text of the source
• cites Thandi et al., 2023 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Toli et al. (2016) reported a large effect (d = 0.99) for implementation intentions in populations with mental health problems.

“At first glance, our effect estimate appears considerably lower than the large effect (d = 0.99) reported by Toli et al. for implementation intentions in populations with mental health problems.”

• cites Toli et al., 2016 sourceSupported

May combine two claims.

These findings support the idea that implementation intentions are a simple strategy that is especially suitable for children and capable of compensating for developmental or clinical self-regulation deficits.

“These findings support the idea that implementation intentions are a ‘simple’ strategy (Gollwitzer, 1999) that is especially suitable for children and capable of compensating for developmental or clinical self‐regulation deficits.”

• cites Gollwitzer, 1999 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Research finds high levels of problematic screen time and social media use in children and young adolescents.

“For example, research finds high levels of problematic screen time and social media use in children and young adolescents (Nagata et al., 2025), yet none of the studies in this meta‐analysis addressed this challenge.”

• cites Nagata et al., 2025 sourceSupported

May combine two claims.

Plan creation was not significant in a two-sided test; the tendency was for experimenter-provided plans to be more effective than self-created plans (b = 0.44, p = 0.98).

“The moderator plan creation was not significant in a two‐sided test; the tendency was for experimenter‐provided plans to be more effective than self‐created plans (b = 0.44, p = 0.98).”

May combine two claims.

When influential cases were removed, type of sample and number of plans also showed moderating effects.

“When influential cases were removed, type of sample and number of plans also showed moderating effects.”

May combine two claims.

The within-person studies had small effects of 0.08, 0.11 and 0.28, with an overall g = 0.12, 95% CI [0.02; 0.21], p = .020, k = 3.

“The effects in the within‐person studies were rather small, namely, 0.08, 0.11 and 0.28 (overall g = 0.12, 95% CI [0.02; 0.21], p = .020, k = 3).”

The quote only partly states this claim.

The average pooled effect across all iterations using a random sample of 10% of all possible combinations of alternative outcome variables was g = 0.30 (SD = 0.02, range: 0.24-0.36, average I 2 = 60.26%).

“The average pooled effect across all iterations was g = 0.30 (SD = 0.02, range: 0.24–0.36, average I 2 = 60.26%).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Alternative effect sizes based on statistical tests that controlled for other variables could be calculated for nine out of the 52 effect sizes.

“Such alternative effect sizes could be calculated for nine out of the 52 effect sizes.”

May combine two claims.

The predictors fit of the plan to the outcome, number of plans, setting (laboratory vs. field) and time between plan creation and first outcome assessment had possible variable importance in the analysis excluding influential cases but did not reach significance in the meta-regression.

“The predictors fit of the plan to the outcome, number of plans, setting (laboratory vs. field) and time between plan creation and first outcome assessment also had possible variable importance in this analysis but did not reach significance in the meta‐regression.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Some included studies explicitly stated that they were interested in the intention-to-treat effect.

“Some studies (e.g. Armitage & Sprigg, 2010; Conner & Higgins, 2010) explicitly stated that they were interested in the intention‐to‐treat effect, that is, the effect of assigning children to the implementation intention intervention regardless of their adherence to the intervention.”

May combine two claims.

Thirty-three studies used implementation intentions that were predetermined by the experimenter; only 14 studies asked children to formulate their own plans, and five studies let children create plans by choosing from provided examples.

“Thirty‐three studies used implementation intentions that were predetermined by the experimenter; only 14 studies asked children to formulate their own plans, and five studies let children create plans by choosing from provided examples.”

May combine two claims.

The three-parameter selection model estimated an effect size of 0.38 (p < .001), with substantial heterogeneity among the studies (LRT(1) = 20.95, p < .001).

“The estimated effect size was 0.38 (p < .001), with substantial heterogeneity among the studies (LRT(1) = 20.95, p < .001).”

May combine two claims.

Crăciun and Băban (2008) had a comparatively large sample size and thus influence on the overall result.

“This study had a comparatively large sample size and thus influence on the overall result.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Ten out of the 52 effect sizes could be re-calculated based on later outcome assessment.

“Ten out of the 52 effect sizes could be re‐calculated based on later outcome assessment.”

Probably not a substantial claim: may not pass S3 (load-bearing).

There were 40 effect sizes from non-clinical samples.

“Taken together, there were 12 effect sizes from clinical samples (11 ADHD, 1 repaired cleft lip and/or palate) and 40 from non‐clinical samples.”

May combine two claims.

For the effect of assignment to the intervention, there were 29 effect sizes with low or some concerns, and the pooled effect size across these studies was g = 0.47 (95% CI [0.31; 0.64], k = 29; Q(28) = 107.32 with p < .001; I 2 = 73.9%).

“For the effect of assignment to the intervention, there were 29 effect sizes with low or some concerns. The pooled effect size across these studies was g = 0.47 (95% CI [0.31; 0.64], k = 29; Q(28) = 107.32 with p < .001; I 2 = 73.9%).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Alternative outcome variables were identified for 18 of the 40 effect sizes.

“Such alternative outcome variables were identified for 18 of the 40 effect sizes.”

May combine two claims.

The funnel plot revealed asymmetry, with three effect sizes showing high standard errors and strong statistical significance of positive effects, while more precise estimates tended to cluster on the left side of the pooled effect.

“The plot revealed asymmetry, with three effect sizes showing high standard errors and strong statistical significance of positive effects. In contrast, more precise estimates tended to cluster on the left side of the pooled effect.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The mean age of participants was 10.67 years (range: 4.5-12.99 years).

“Participants were children, with a mean age of 10.67 years (range: 4.5–12.99 years).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Forty of the 52 effect sizes are based on non-clinical samples.

“Furthermore, 40 of the 52 effect sizes are based on non‐clinical samples.”

Probably not a substantial claim: may not pass S3 (load-bearing).

There were 12 effect sizes from clinical samples (11 ADHD, 1 repaired cleft lip and/or palate).

“Taken together, there were 12 effect sizes from clinical samples (11 ADHD, 1 repaired cleft lip and/or palate) and 40 from non‐clinical samples.”

May combine two claims.

For the per-protocol analysis, no effect size was rated as low risk of bias, 19 had some concerns and 33 were considered of high concern, primarily due to insufficient reporting of whether children created implementation intentions of sufficient quality or reported problems with the quality of child-created plans.

“With regard to the per‐protocol analysis, risk of bias was generally higher: No effect size was rated as low risk of bias, 19 had some concerns and 33 were considered of high concern, primarily due to insufficient reporting of whether children created implementation intentions of sufficient quality, or due to reported problems with the quality of child‐created plans (e.g. Beall, 2011; Schaaf et al”

Probably not a substantial claim: may not pass S3 (load-bearing).

In two further studies, separate effect sizes were computed for younger and older children.

“In two further studies, separate effect sizes were computed for younger and older children (Kretschmer‐Trendowicz et al., 2021, Study 1 and Study 2).”

• cites Kretschmer‐Trendowicz et al., 2021 sourceNot checked — no full text of the source

May combine two claims.

Only three of the 52 effect sizes stem from within-person comparisons, so no meaningful quantitative comparison was possible.

“Only three of the 52 effect sizes stem from within‐person comparisons; thus, no meaningful quantitative comparison was possible.”

May combine two claims.

Ratings of some concerns were mainly due to missing preregistration of outcomes and analysis plans, while high risk of bias ratings were often based on baseline differences between groups, potentially selective dropout rates, or exclusion of participants after randomization.

“Ratings of some concerns were mainly due to missing preregistration of outcomes and analyses plans, while ratings of high risk of bias were often based on baseline differences between groups, potentially selective dropout rates, or the exclusion of participants after randomization.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The large majority of studies used if and then components that described external rather than internal cues and behaviours.

“For example, the large majority of studies used if and then components that described external (rather than internal) cues and behaviours.”

May combine two claims.

For the intention-to-treat analysis, two of the 52 effect sizes were rated as low risk of bias, 27 had some concerns and 23 were judged as high risk of bias.

“Regarding the intention‐to‐treat analysis, two of the 52 effect sizes were rated as low risk of bias, 27 had some concerns and 23 were judged as high risk of bias.”

Probably not a substantial claim: may not pass S3 (load-bearing).

When replacing all 10 of these effect sizes at once, the average time between intervention and assessment changed to 88.25 days (SD = 286.59 days, range: 0-1461 days).

“When replacing all 10 of these effect sizes at once, the average time between intervention and assessment changed to 88.25 days (SD = 286.59 days, range: 0–1461 days).”

May combine two claims.

For the effect of adhering to the intervention, there were 19 effect sizes with low or some concerns, and the pooled effect size across these studies was g = 0.52 (95% CI [0.24; 0.80], k = 19; Q(18) = 62.02 with p < .001; I 2 = 71.0%).

“For the effect of adhering to the intervention, there were 19 effect sizes with low or some concerns. The pooled effect size across these studies was g = 0.52 (95% CI [0.24; 0.80], k = 19; Q(18) = 62.02 with p < .001; I 2 = 71.0%).”

May combine two claims.

Crăciun and Băban (2008) had an effect size of 0.85, considerably larger than the pooled effect.

“Although not the second largest, the effect size of 0.85 in the study by Crăciun and Băban (2008) was also considerably larger than the pooled effect.”

May combine two claims.

The cross-validated explained variance was R 2 cv = 0.40 and the out-of-bag explained variance was R 2 oob = 0.01.

“The cross‐validated explained variance was R 2 cv = 0.40; the out‐of‐bag explained variance was R 2 oob = 0.01.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The average time between intervention and the first outcome assessment was 67.04 days (SD = 283.11 days, range: 0-1461 days).

“The average time between intervention and the first outcome assessment was 67.04 days (SD = 283.11 days, range: 0–1461 days).”

Probably not a substantial claim: may not pass S3 (load-bearing).

The included studies were published between 1975 and 2025.

“The studies were published between 1975 and 2025 (median publication year = 2011).”

May combine two claims.

Most studies do not publish their materials, and descriptions of interventions often lack details on how the plans were created or provided.

“Most studies do not publish their materials, and descriptions of interventions often lack details on how the plans were created or provided.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Studies used different strategies to ensure the plan's accessibility.

“In addition, studies used different strategies to ensure the plan's accessibility.”

Classified as probably not a substantial claim.

There was a large variety of instructions and materials across studies.

“The large variety of instructions and materials we encountered across studies highlights the need for evidence‐based guidelines on designing age‐appropriate materials that support the cognitive processes involved in planning.”

May combine two claims.

Just over half of the included studies assessed motivation in some form, but motivation was hardly ever tested as a moderator.

“While just over half of the included studies assessed motivation in some form, it was hardly ever tested as a moderator.”

Probably not a substantial claim: may not pass S3 (load-bearing).

We found earlier studies that used comparable if-then planning strategies with children and were cited in foundational implementation intention research.

“Although the concept of implementation intentions is typically attributed to Gollwitzer (first published in the 1990s), we found earlier studies that used comparable if‐then planning strategies with children and were cited in foundational implementation intention research (e.g. Wieber et al., 2011).”

May combine two claims.

When excluding influential cases, interventions showed stronger effects in clinical samples, which consisted almost exclusively of children with ADHD.

“Furthermore, when excluding influential cases, interventions showed stronger effects in clinical samples, which consisted almost exclusively of children with ADHD.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

In most cases, children had to write down the plan themselves either digitally or on paper, but especially in preschool children, plans were provided orally.

“In most cases, children had to write down the plan themselves either digitally (e.g. Breitwieser et al., 2023) or on paper (e.g. Beall, 2011; Crăciun & Băban, 2008). However, especially in preschool children, plans were provided orally (Mischel & Patterson, 1976; Patterson & Mischel, 1975, 1976).”

May combine two claims.

Only 14 studies involved children generating their own plans; in the remaining 38 studies, plans or examples of plans were provided by the experimenter.

“Only 14 studies involved children generating their own plans; in the remaining 38 studies, plans or examples of plans were provided by the experimenter.”

May combine two claims.

Higgins and Conner (2003) aimed to reduce smoking initiation in adolescents, but only three out of 53 adolescents in the control group began smoking, suggesting that the goal was too easy to be influenced by the intervention.

“For instance, Higgins and Conner (2003) aimed to reduce smoking initiation in adolescents, but only three out of 53 adolescents in the control group began smoking, suggesting that the goal was too easy to be influenced by the intervention.”

May combine two claims.

Sheeran et al.'s inclusion period spanned 1994-2019, while we included 14 effect sizes published after 2019 as well as four effect sizes from studies conducted in the 1970s.

“Furthermore, Sheeran et al.'s inclusion period spanned 1994–2019, while we included 14 effect sizes published after 2019 as well as four effect sizes from studies conducted in the 1970s by Walter Mischel and Charlotte Patterson.”

May combine two claims.

The variable of experimenter-provided versus self-created plans emerged as relevant in the MetaForest analysis after excluding influential cases but did not quite reach significance.

“This variable emerged as relevant in the metaforest analysis (after excluding influential cases) but did not quite reach significance.”

May combine two claims.

Only one study used an internal cue, preventing any conclusions about the comparison between external and internal cues.

“However, only one study used an internal cue (Mau et al., 2019, Study 2), thus preventing any conclusions about this comparison.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Usually, children were asked to repeat their plan three times, but some studies asked children to repeat the plan only once, and others just instructed children to memorize the plan.

“Usually, children were asked to repeat their plan three times (e.g. Bleize et al., 2022; Gawrilow et al., 2011a; Gawrilow et al., 2011b; Kretschmer‐Trendowicz et al., 2021; Mau et al., 2019; Zimmermann & Meier, 2010). However, some studies asked the children to repeat the plan only once (e.g. Duckworth et al., 2013; Yang et al., 2023), and others just instructed children to memorize the plan (e.g.”

Classified as probably not a substantial claim.

Differences in instructions and materials are difficult to capture, especially because many studies do not report the procedure with sufficient detail.

“Furthermore, differences in instructions and materials, as mentioned above, are difficult to capture, especially since many studies do not report the procedure with sufficient detail.”

May combine two claims.

Our findings revealed stronger effects in studies with younger children and children with ADHD.

“Our findings revealed stronger effects in studies with younger children and children with ADHD.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

School entry has an impact on the development of self-regulation, with research suggesting that schooling can improve executive functions.

“School entry also has an impact on the development of self‐regulation, with research suggesting that schooling can improve executive functions (McCrea et al., 1999; see also Morrison et al., 2010).”

• cites McCrea et al., 1999 sourceNot checked — no full text of the source
• cites Morrison et al., 2010 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

During early childhood, the development of self-regulation is closely linked to brain development.

“During early childhood, the development of self‐regulation is closely linked to brain development (Diamond, 2002; McClelland et al., 2018).”

• cites Diamond, 2002 sourceNot checked — no full text of the source
• cites McClelland et al., 2018Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Self-regulation abilities in childhood are important for long-term academic achievement and healthy development.

“Self‐regulation abilities in childhood are important for long‐term academic achievement and healthy development.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The development of self-regulation in early childhood may pave the way for success later in life.

“Thus, the development of self‐regulation in early childhood may pave the way for success later in life.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Both authors independently screened full texts for final inclusion; disagreements were resolved through discussion.

“Subsequently, both authors independently screened the full texts of the remaining studies for the final inclusion; disagreements were resolved through discussion.”

Classified as probably not a substantial claim.

Implementation intentions have been proposed as a volitional strategy that can help bridge the intention-behaviour gap.

“Since intentions explain less than 30% of the variance in behaviour across behavioural domains (Sheeran, 2002), implementation intentions have been proposed as a volitional strategy that can help bridge this gap.”

• cites Sheeran, 2002 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

As cognitive functions develop during childhood and adolescence, self-regulation interventions need to be simple and easy to use.

“As cognitive functions develop during childhood and adolescence, self‐regulation interventions need to be simple and easy to use.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Moderator analyses are considered exploratory.

“Moderator analyses are thus considered exploratory.”

Probably not a substantial claim: may not pass S2 (checkable).

Identifying moderators that explain this heterogeneity is important for designing more effective interventions and uncovering the boundary conditions of implementation intentions in children.

“Identifying moderators that explain this heterogeneity is important for designing more effective interventions and uncovering the boundary conditions of implementation intentions in children.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Each of the two raters independently entered descriptive information and effect size information in coding sheets.

“For the final data extraction from the eligible studies, each of the two raters independently entered the descriptive information about the studies and the information needed to calculate effect sizes in the coding sheets (see S5 and S6 on the OSF project page).”

Classified as probably not a substantial claim.

The overall effectiveness of implementation intentions may differ between children and adults due to developmental differences in self-regulation abilities.

“Furthermore, the overall effectiveness of implementation intentions may differ between children and adults due to developmental differences in self‐regulation abilities.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Disagreements in data extraction were resolved by discussion.

“We resolved disagreements by discussion.”

May combine two claims.

The observed discovery rate (ODR = 0.35, 95% CI [0.22, 0.49]) exceeded the expected discovery rate (EDR = 0.23, 95% CI [0.05, 0.68]), indicating that statistically significant findings are likely overrepresented relative to what would be expected based on the estimated effect-size distribution and statistical power.

“The observed discovery rate (ODR = 0.35, 95% CI [0.22, 0.49]) exceeded the expected discovery rate (EDR = 0.23, 95% CI [0.05, 0.68]), indicating that statistically significant findings are likely overrepresented relative to what would be expected based on the estimated effect‐size distribution and statistical power”

Probably not a substantial claim: may not pass S3 (load-bearing).

p-curve and p-uniform* are generally not considered robust when between-study heterogeneity is higher than I 2 = 50%.

“Note, however, p‐curve and p‐uniform* are generally not considered robust when the between‐study heterogeneity is higher than I 2 = 50% (van Aert & van Assen, 2021).”

• cites van Aert & van Assen, 2021Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

A systematic literature search was conducted in PsycInfo, PSYNDEX, and PubMed in February 2024.

“A systematic literature search was conducted in the following bibliographic databases in February 2024: PsycInfo, PSYNDEX (database of psychological publications from the German‐speaking countries) and PubMed.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Adolescence marks another sensitive period of self-regulation development when adult-like self-regulatory strategies emerge.

“Adolescence marks another sensitive period of self‐regulation development when adult‐like self‐regulatory strategies emerge (McClelland et al., 2018).”

• cites McClelland et al., 2018Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Several sensitivity analyses were conducted to test the robustness of the meta-analytic results.

“We conducted several sensitivity analyses to test the robustness of the meta‐analytic results.”

May combine two claims.

Higgins and Conner (2003) did not find a primary effect, but only effects involving moderators based on the Theory of Planned Behavior.

“Armitage and Sprigg (2010) observed an effect of implementation intentions, whereas Higgins and Conner (2003) did not find a primary effect, but only effects involving moderators based on the Theory of Planned Behavior.”

• cites Higgins and Conner (2003) sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

To validate the search strategy, we verified that all 16 articles were found.

“To validate the search strategy, we verified that all 16 articles were found.”

Probably not a substantial claim: may not pass S3 (load-bearing).

An additional search was conducted in ProQuest Dissertations & Theses Global to identify unpublished doctoral dissertations.

“To further identify unpublished doctoral dissertations, an additional search was conducted in ProQuest Dissertations & Theses Global.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Both authors independently screened titles and abstracts for eligibility.

“Both authors independently screened the titles and abstracts of identified study records for eligibility.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Identity development and schooling lead to a shift in goal focus towards more long-term and outcome-related goals.

“In addition, identity development and schooling lead to a shift in goal focus towards more long‐term and outcome‐related goals (Moersdorf et al., 2022; Wigfield et al., 2011).”

• cites Moersdorf et al., 2022 sourceNot checked — no full text of the source
• cites Wigfield et al., 2011Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Especially younger children may lack these prerequisites as well.

“Especially younger children may lack these prerequisites as well.”

Probably not a substantial claim: may not pass S3 (load-bearing).

A recent meta-analysis of meta-analyses revealed a combined effect size of d = 0.54, 95% CI [0.51, 0.57], range: 0.24-0.99.

“A recent meta‐analysis of meta‐analyses revealed a combined effect size of d = 0.54, 95% CI [0.51, 0.57], range: 0.24–0.99 (Keller et al., 2020).”

• cites Keller et al., 2020 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The risk of reporting bias in the meta-analytic result was assessed.

“To assess the risk of reporting bias in the meta‐analytic result, the following analyses have been performed.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The coding sheets were piloted and optimized by coding 10 studies on implementation intention effects in adults.

“The coding sheets were piloted and optimized by coding a total of 10 studies on implementation intention effects in adults (Achtziger et al., 2008; Adriaanse et al., 2009; de Nooijer et al., 2006; de Vet et al., 2009; Latimer et al., 2006; Luszczynska, 2006; Prestwich et al., 2009; Prestwich et al., 2010; Rodrigues et al., 2013; Smith et al., 2022).”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

This great heterogeneity in materials and instructions might in part reflect different study goals.

“This great heterogeneity in materials and instructions might in part reflect different study goals.”

May combine two claims.

Sheeran et al. (2024) included 642 effect sizes overall, 31 of which came from samples of children under the age of 12 and yielded an average effect size of d = 0.43 (95% CI [0.27, 0.59]).

“Their analysis included 642 effect sizes overall, 31 of which came from samples of children under the age of 12 that yielded an average effect size of d = 0.43 (95% CI [0.27, 0.59]).”

• cites Sheeran et al., 2024Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

The behaviour-goal contingency may be weak if the specified behaviour has limited impact on the goal.

“Second, the behaviour‐goal contingency may be weak if the specified behaviour has limited impact on the goal.”

May combine two claims.

When applying robust estimation methods, Sheeran et al.'s overall effect size dropped from 0.36 to 0.15; they did not provide a separate corrected estimate for child samples.

“When applying robust estimation methods, their overall effect size dropped from 0.36 to 0.15; however, they did not provide a separate corrected estimate for child samples.”

• cites Sheeran et al., 2024Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Taken together, these findings suggest that, while the effects of implementation intentions in children are robust, they could be somewhat smaller than previously assumed.

“Taken together, these findings suggest that, while the effects of implementation intentions in children are robust, they could be somewhat smaller than previously assumed.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

Strategies such as visibly displaying the plan in the classroom, attaching it to a desk or embedding it into daily routines may enhance accessibility and support plan execution.

“Strategies such as visibly displaying the plan in the classroom, attaching it to a desk or embedding it into daily routines may enhance accessibility and support plan execution.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Visual formats may be more effective than verbal-only formats for helping children encode and represent the plan in memory.

“Visual formats (e.g. drawings) may be more effective than verbal‐only formats for helping children encode and represent the plan in memory.”

• AssumptionThe authors state this as an assumption: Representations develop from more modal (image-like) to more amodal (abstract).

Classified as probably not a substantial claim.

When the measured outcome is too broad or distal from the specific behaviour targeted in the plan, intervention effects may be obscured.

“When the measured outcome is too broad or distal from the specific behaviour targeted in the plan, intervention effects may be obscured.”

May combine two claims.

Although our pooled effect size falls within the 95% confidence interval reported by Sheeran et al. for child samples, it represents a more conservative estimate.

“Although our pooled effect size falls within the 95% confidence interval reported by Sheeran et al. for child samples, it represents a more conservative estimate.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The plan-behaviour contingency may fail if the situational cue is too vague to facilitate automatic action initiation.

“First, the plan–behaviour contingency may fail, for instance, if the situational cue is too vague to facilitate automatic action initiation.”

May combine two claims.

While implementation intentions are effective when provided to children, it is unclear how best to support children in generating high-quality plans themselves.

“This suggests that, while implementation intentions are effective when provided to children, it is unclear how best to support children in generating high‐quality plans themselves.”

May combine two claims.

When studies with a high risk of bias were excluded, the effect size increased to g = 0.52, indicating that high-quality studies achieved larger effect sizes.

“Conversely, when studies with a high risk of bias were excluded, the effect size increased to g = 0.52, indicating that high‐quality studies achieved larger effect sizes.”

Classified as probably not a substantial claim. May combine two claims.

According to Implementation Intention Theory, the effectiveness of implementation intentions depends on the strength of the association between the if and then components of the plan as well as the plan's accessibility.

“According to Implementation Intention Theory, the effectiveness of implementation intentions depends on the strength of the association between the ‘if’ and ‘then’ components of the plan as well as the plan's accessibility (Achtziger & Gollwitzer, 2018; Webb & Sheeran, 2008).”

• cites Achtziger & Gollwitzer, 2018 sourceNot checked — no full text of the source
• cites Webb & Sheeran, 2008 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Mental contrasting is a strategy designed to enhance goal commitment when expectations of success are high.

“a strategy designed to enhance goal commitment when expectations of success are high (Oettingen et al., 2001).”

• cites Oettingen et al., 2001 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Sheeran et al. report strong evidence of publication bias in their meta-analysis.

“Furthermore, Sheeran et al. report strong evidence of publication bias in their meta‐analysis.”

• cites Sheeran et al., 2024Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Implementation intentions appear to be beneficial even for relatively young children and for those with pronounced self-regulatory difficulties.

“Importantly, they appear to be beneficial even for relatively young children and for those with pronounced self‐regulatory difficulties.”

May combine two claims.

Pre-structured sheets emerged as a significant moderator when the two influential studies were excluded, offering tentative evidence that guided support may improve the quality and effectiveness of self-generated plans.

“Pre‐structured sheets emerged as a significant moderator when the two influential studies were excluded, offering tentative evidence that guided support may improve the quality and effectiveness of self‐generated plans.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Both action and coping planning are regularly used as part of child cognitive behavioural psychotherapy.

“Similarly, both action and coping planning are regularly used as part of child cognitive behavioural psychotherapy (e.g. Podell et al., 2010).”

• cites Podell et al., 2010 sourceNot checked — no full text of the source

May combine two claims.

Lack of transparency in outcome specification complicates interpretation and increases the risk of post hoc outcome switching or selective reporting.

“This lack of transparency complicates interpretation and increases the risk of post hoc outcome switching or selective reporting (Nosek et al., 2018).”

• cites Nosek et al., 2018 sourceNot checked — no full text of the source

May combine two claims.

In our analysis, we found evidence of small study effects, and the overall effect size decreased when statistical outliers were excluded.

“In our analysis, we found evidence of small study effects, and the overall effect size decreased when statistical outliers were excluded.”

Probably not a substantial claim: may not pass S2 (checkable).

The absence of a significant moderating effect cannot be taken as evidence for the absence of moderation.

“First, the absence of a significant moderating effect cannot be taken as evidence for the absence of moderation.”

Classified as probably not a substantial claim. May combine two claims.

Implementation intention theory emphasizes the importance of forming a strong mental representation of the plan in memory and keeping it accessible.

“implementation intention theory emphasizes the importance of forming a strong mental representation of the plan in memory and keeping it accessible (Achtziger & Gollwitzer, 2018; Webb & Sheeran, 2008).”

• cites Achtziger & Gollwitzer, 2018 sourceNot checked — no full text of the source
• cites Webb & Sheeran, 2008 sourceNot checked — no full text of the source

May combine two claims.

These findings suggest that implementation intentions are particularly beneficial for children with self-regulatory difficulties, whether because they have not yet reached certain developmental milestones or because they have a neurodevelopmental impairment.

“These findings suggest that implementation intentions are particularly beneficial for children with self‐regulatory difficulties, whether because they have not yet reached certain developmental milestones or because they have a neurodevelopmental impairment.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

Practitioners such as paediatricians, child psychotherapists or teachers use plans on a daily basis, often very similar to implementation intentions.

“Practitioners such as paediatricians, child psychotherapists or teachers use plans on a daily basis, often very similar to implementation intentions.”

May combine two claims.

This meta-analysis provides robust evidence that implementation intentions can support children in achieving a range of desirable outcomes, including improvements in academic, health-related and interpersonal behaviours as well as executive functioning.

“This meta‐analysis provides robust evidence that implementation intentions can support children in achieving a range of desirable outcomes, including improvements in academic, health‐related and interpersonal behaviours as well as executive functioning.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Research on frequent plan reminders suggests that children may benefit from support in maintaining and recalling their implementation intentions over time.

“Furthermore, research on frequent plan reminders suggests that children may benefit from support in maintaining and recalling their implementation intentions over time (Breitwieser et al., 2023).”

• cites Breitwieser et al., 2023 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The format in which plans are presented may play a critical role and the optimal format may vary with age.

“This suggests that the format in which plans are presented may play a critical role and that the optimal format may vary with age.”

• AssumptionThe authors state this as an assumption: Representations develop from more modal (image-like) to more amodal (abstract).

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Important goals and life domains are currently underrepresented in the literature, limiting the generalizability of the findings.

“At the same time, it could be argued that there are important goals and life domains that are currently underrepresented in the literature, thus limiting the generalizability of our findings.”

Classified as probably not a substantial claim. May combine two claims.

The substantial heterogeneity in effect sizes highlights the need for a better understanding of the conditions under which implementation intentions are most effective and for whom.

“At the same time, the substantial heterogeneity in effect sizes highlights the need for a better understanding of the conditions under which implementation intentions are most effective and for whom.”

May combine two claims.

Peer relations, changes in family dynamics and social identity become increasingly important during early adolescence.

“Only two studies in this meta‐analysis targeted interpersonal challenges (Bleize et al., 2022; Mauduy et al., 2023), although peer relations, changes in family dynamics and social identity become increasingly important during early adolescence (Steinberg & Morris, 2001).”

• cites Steinberg & Morris, 2001 sourceNot checked — no full text of the source