Skip to main content

Verified by Psychology Today

Child Development

Beyond the Marshmallow Test: Rethinking Delayed Gratification

Response inhibition training does not change children’s brains or behavior.

Key points

  • The ability to delay gratification (a type of cognitive control) is associated with positive outcomes.
  • Can training improve this ability?
  • New research proposes surprising answers.

Cognitive control is a term used to describe the processes that enable us to think, feel, and act in flexible, goal-directed ways. Psychologists have been fascinated by the concept of cognitive control ever since Walter Mischel’s famous marshmallow studies purported to show the benefits of delayed gratification (an aspect of cognitive control) for later development in children.

Measures of childhood cognitive control have since been shown to predict a host of later outcomes, including social skills, academic performance, and mental health. Cognitive control develops gradually during childhood, enabled by brain maturation and the child’s growing experience.

The questions for psychologists, parents, and policymakers are:

  1. Can cognitive control be speeded up through intentional training?
  2. Can such training improve behavioral outcomes in children?
RonaldPlett for Pixabay
Source: RonaldPlett for Pixabay

Until recently, the answers to these questions appeared to be yes. Cognitive control has been viewed as a capacity or skill. As such, went the argument, children can be trained to become better at it, which in turn will result in better developmental outcomes. Consequentially, many training programs—and studies looking at them—have been developed.

To be effective, cognitive control training must demonstrate two outcomes. The first is “near transfer”: how well kids perform on their training task. The second is “far transfer,” which is more important because it measures how the training generalizes to real-life behavior and health outcomes.

Childhood interventions are notoriously costly in terms of time and money. Thus, psychologists (and policymakers) face the urgent task of teasing apart effective from non-effective ones. Unfortunately, much of the work to date on cognitive control training has been plagued by methodological problems, and the results have been less than clear and often contradictory.

Recently (2024), British researcher Keertana Ganesan and colleagues sought to provide some definitive answers by conducting a high-quality, preregistered experiment involving an active control group, clearly defined training mechanisms, a large sample of participants, and a comprehensive set of outcome measures taken multiple times. The researchers looked at whether response inhibition can be improved by training, and whether such improvements will lead to changes in multiple behavioral and neural (brain development) outcomes.

In the study, 235 children ages 6-13 were randomly assigned to an experimental or control group. The groups were matched for gender, age, school, and class. During the structured 8-week intervention, both groups were trained in a similar gamified task under a similar protocol with one difference: The experimental group participants were trained to inhibit (delay) their responses while the control group participants were trained to increase response speed.

The researchers then compared the groups on a host of measures shown to relate to cognitive control and, specifically, response inhibition, including social and interpersonal decision-making, academic achievement, fluid reasoning, mental health (internalizing and externalizing symptoms), as well as various neural indices of brain function, structure, and connectivity. They repeated these measures at a 1-year follow-up.

Results first showed that the training worked, producing “long-lasting improvements of closely related measures of cognitive control at the 1-year follow-up.” In other words, in terms of “near transfer,” both groups became—and remained—better at the tasks they were trained to master.

However, when looking at the “far transfer” results, the authors found that the training “had no impact on any behavioral outcomes (decision-making, academic achievement, mental health, fluid reasoning, and creativity) or neural outcomes (task-dependent and intrinsic brain function and gray and white matter structure).” They conclude: “Targeted training of response inhibition does little to change children’s brains or their behavior."

The authors set out to test three further hypotheses regarding cognitive control training. First, they tested whether “near transfer” effects lead to "far transfer" effects. In other words, does success at near transfer predict success at far transfer?

Second, they tested whether “far transfer” effects that fail to emerge immediately may emerge over time. Finally, making use of the occurrence of COVID-19 during their experiment, they tested whether cognitive control might buffer against the onset of mental health problems.

The evidence failed to support all three hypotheses. “There was... no indication of training effects emerging over time, nor did the presence of near transfer effects mediate the likelihood of far transfer. Finally, training response inhibition did not act as a buffer to mental health problems as a result of major social stressors, such as COVID-19.”

Their conclusion: “In sum, response inhibition training appears to do little to alter children’s brains or their behavior in long-lasting ways… Given the considerable policy implications of how children can be supported in their development, these findings caution against any further investment in seeking to improve response inhibition specifically and cognitive control more generally through trainings that canonically aim to boost these capacities wholesale.”

In other words, the hope and hype regarding cognitive control training are not backed by evidence.

Although the study is not without limitations (short training duration, non-representative affluent sample), it joins a growing body of evidence suggesting that the old model of cognitive control as a limited capacity is incorrect. An alternative emerging view holds that “cognitive control is assigned a value as a function of subjectively perceived effort and the likely reward or goal priority.” As such, "it could be improved in ways that lead to changes in other domains by targeting motivation and effort expenditure.”

In other words, cognitive control may be a byproduct of motivational factors. Children may learn to improve their cognitive control if given training tasks they perceive as reliable means of achieving worthy rewards that justify the effort.

Research on this hypothesis is underway. In science, the gratification of definitive answers is often delayed...

More from Noam Shpancer Ph.D.
More from Psychology Today