Walter Mischel's Marshmallow Experiment

In the world of psychology, researchers often base their works off of previous experiments that revolutionized their fields and dramatically changed the way of thinking during their time. One example of these revolutionary experiments is Walter Mischel’s 1966 Marshmallow Test, which provided information regarding self-control amongst children exposed to delay-of-gratification. During the late 1960’s when the field of social psychology was in a state of transition between methods of experimentation, the Marshmallow Test was one of the original tests that first used independent and dependent variables. These early experiments set the premise for later experiments to be performed after the definition of experimentation was changed. With this knowledge,

the history of social psychology, the Marshmallow Test, and recent, derivative experiments will be discussed in further detail.
Through a series of tests involving young children, Mischel and his colleagues set out to determine the role of self-regulation by exposing them to delay of gratification.

The participants were separated into two groups: one with the reward(s) in front of them while waiting and the other without the presence of the reward. Both groups were told if they waited 15 minutes, they would receive two marshmallows, but if they did not wait the full amount of time, they would only receive one. In Mischel & Ebbesen’s (1970) “Attention in delay of gratification,” they hypothesized that as the degree of attention paid to the delayed rewards increased, the length of time which the children waited would increase. To their surprise as well as others, their results did not support their hypothesis, and the thoughts regarding self-regulation of the time changed significantly (Mischel & Ebbesen,

1970).
Mischel originally believed that self-regulation is controlled by “hot” and “cool” systems that stem from the brain. The areas most commonly referred to as the “hot” areas of the brain include the subcortical and limbic system structures, whereas the “cool” areas consist of the prefrontal areas. Both of these systems drive the processes involved in choice and decision making. The “hot” system, for example, leads us to make impulsive decisions, but the “cool” system is there to balance this impulsivity. Mischel thought that using the “cool” system to neutralize the “hot” system would be a rational method of controlling this instinctual behavior (Kurti, 2015), but this idea was one that might not have been correct.
One opposing belief to the “hot-cold systems” as described by Mischel was Sigmund Freud’s explanation of the transition from primary to secondary processes (Mischel & Ebbesen, 1970).

The primary process is “the baby’s act of creating in its mind an image of an object that will fulfill a need.” This process is controlled by what Freud terms the id, which refers to the natural, unconscious instincts and drives that a person is born with. The secondary process, however, converts these images into action and is controlled by the ego (Cordón, 2015). Thus, the transition from the unconscious to the conscious opposes Mischel’s belief of how delay of gratification occurs (Mischel & Ebbesen,

1970).
These opposing beliefs were not uncommon during the late 1960s and early 1970s due to the instability of the field of social psychology of the time. Talk of experiments within social psychology first began in the 1920s and expanded into the 1930s, a time that initiated the use of independent and dependent variables. Throughout the 1940s, more people started to create experiments using these variables, and it wasn’t until the 1960s that the concept of experimentation was fully established. That being said, the Marshmallow Test was performed during a period of transition. One of the main contributions to the instability of social psychology was Kuhn’s (1970) definition of a paradigm, which is defined as “a framework of assumptions.” Many people within the field did not fully understand this new term, but it led to the conviction that social psychologists should be using a laboratory setting in order to collect data (Parker, 2013). Therefore, Mischel and Ebbesen were some of the first to use this type of experimentation successfully.
The history of social psychology has paved the way for researchers to expand their knowledge by using the design of past experiments to further research concepts. One example of a present interpretation of the Marshmallow Test is Bucciol et al.’s (2011) “Temptation and productivity: A field experiment with children.” Within this experiment, they hypothesized that willpower depletion reduces productivity and used this hypothesis to test the concept of self-regulation. Bucciol et al. (2011) explains,
The key feature of the Mischel and Ebbesen design is that children were shown a desirable food reward but told they could not consume it unless they successfully completed some task (waiting). Our design is similar, with an exception that instead of ‘waiting’ our task is to play a game, the reward for which is the tempting food (Bucciol et al., 2011).
Using children between the ages of six and thirteen gave the researchers a range that would allow them to accurately test variation amongst different changes groups. Mischel & Ebbesen (1970), however, did not have as wide of a range amongst their participants; they studied nursery children between the ages of 3 years, 6 months and 5 years, 8 months. Their decision to use children in these low age groups is one that is risky because they have only just finished development of self-control at age 3-4. Hence, Bucciol et al.’s (2011) results might have differed slightly from Mischel & Ebbesen’s (1970) results.
Another recent study related to the concept of self-regulation is Moffitt et al.’s (2011) “A gradient of childhood self-control predicts health, wealth, and public safety.” This experiment included over one thousand participants who lived in Dunedin, New Zealand and were followed for thirty years of their lives. Longitudinal studies, such as this, allow researchers to assess development and control variables for each individual; yet, the experiment is very costly and time-consuming. Also, researchers may face problems with the participants’ commitment to the study; often within longitudinal studies, the starting amount of participants is greater than the latter because of people moving for various reasons (i.e. jobs, weather, education, etc.). Another problem that these researchers might have faced is generalizability issues. It is incredibly difficult to find much of a contrast within subjects from all the same cohort, especially when testing self-control (Krause & Corts, 2012). Despite these issues that might have occurred, Moffit et al. (2011) did make a contribution to the field. They provided evidence for the need of interventions in early stages of life to prevent health, financial, and public safety issues later in life.
Intervention to improve self-control has been studied both recently and in the past. One article from the past that describes this is Bugental et al.’s (1978) “Attributional and Behavioral Changes Following Two Behavior Management Interventions with Hyperactive Boys: A Follow-Up Study.” The original study was conducted in 1977, and their results suggested that there was little to no short-term effects. Therefore, they wanted to conduct another study six months later to see if their original study produced any long-term effects instead. Like the original study, they separated a sample of children into two groups based on the type of intervention they received (self-control or social reinforcement) and used two different methods of tutoring to apply the intervention. In order to keep the teachers blind, Bugental et al. (1978) explains,
Teachers making follow-up ratings on the Conners were not the same as those making
initial ratings. They were blind as to what kind of intervention the child had received
but were aware that the child had participated in a tutoring project, and they were also aware of the child's medication status (248).
This allowed the researchers to perform the experiment without any prior biases. Because of this, the teachers were able to successfully able to measure and rank the children accurately. They concluded that the children who received the self-control intervention had an increase in the amount of self-control possessed in the long-term. Thus, the combined effect of Bugental et al.’s (1978) and Moffit et al.’s (2011) articles show that interventions have been and will consistently be used throughout time.
Interventions are a product of social psychology that directly show the contribution that the Marshmallow Test had on the field. If it weren’t for the early experiments of the 1970s such as Mischel & Ebbesen’s (1970) “Attention in delay of gratification,” researchers today might not have had a good foundation on which to build their work. Some of these current experiments vary slightly in the method of experimentation used; however, the transition at the mark of the late 1960s provided a paradigm within the field that is still being used. It is with this information that critiques and further research on self-control can be explored, and the journey through social psychology will prevail.