Personality Psychology (740)

         a blog


Watts et al., (2018)

Filed under: Uncategorized — drcb @ 12:58 pm
32 Comments to “Watts et al., (2018)”
  1. Katherine Chang says:

    I quite enjoyed reading this article as I’ve had my doubts about the Marshmallow Test since learning about it (mainly via pop culture). In fact, a fictional adult-variation of it was used in the movie, The Five-Year Engagement, which is evidence of its popularity and “pizazz”; it’s a very attractive study. Honestly, I can understand why the conclusion might be appealing and why it may not have been questioned for almost forty years – it suggests a simple correlation between waiting and “self-control” and future achievement, which seems to be a very logical connection. However, as stated by the authors, there may be bias with just accepting the results as is because the original study failed to consider other confounding factors that could contribute to achievement. So, I was happy to read that Watts, Duncan, and Quan’s research expanded on the concept of the marshmallow study, including many controls and ultimately challenged the original findings of Shoda et al. (1990); I felt like they addressed many of the reservations that I had about the conclusions that Shoda et al. made.
    Overall, I feel like the authors really took care to examine the variables they added and I don’t have much of a comment or challenge to their methodology. The one thing I have a question on, however, is why did they choose to reduce the waiting time to 7 minutes? Ultimately, it seems like it didn’t matter regarding the significance of the results but it seems like they were unable to draw some further conclusions because such a large proportion made it past the 7-minute ceiling. Additionally, there is a small part of me that wonders if a child is prompted with a 7-minute goal versus a 20-minute goal, would that change their perception of how hard the challenge would be? I’m not sure if this really matters in the end but I found myself perseverating on, “why 7 minutes? Especially if it’s such a limitation.”
    I thought it was interesting (and also appreciated) the authors challenge to what construct the marshmallow test was actually measuring—it had lower correlations with the behavioral measures of attention, impulsivity, or self-control than the measure for mathematical problem solving on the WJ-R. My takeaway from their results and discussion was self-control, future achievement, and probably other constructs are complex and drawing overly simple conclusions might be biased and somewhat misleading. This article has also given me an appreciation for linear regression (since we’ve been learning about this is statistics) as a tool to determine the variability and importance of multiple variables. My last thought is on why the original conclusions might be appealing to the public—I feel like our society is very drawn to success and trying to determine what can predict future success. Because of this, I think some people are very prone to taking results without critically thinking about validity or potential confounds or even causality versus correlation. It reminds me of the silly example people use to demonstrate correlation—when ice cream sales rise, so do homicides. If that is all that is reported, people might jump to the conclusion that ice cream sales causes homicides, which is what I feel is what happened when the original work on the marshmallow study was published—people jumped on the opportunity to create interventions to address self-control and delayed gratification. While these aspects are probably important regardless, it is misleading if they’re based on biased research. The consequence is potentially a family spending a lot of money on these interventions and not receiving the result they expected, potentially at the expense of a child. I think articles like this one are important to pull on the reigns a little in order to try to understand what is actually going on and better inform interventions if the ultimate goal is to create success.

    • Sunghee Kim says:

      Hello, Katherine. Thanks for sharing your thought about the study. I did feel similar as you did while I was reading the article. I did not enjoy reading this article. I really couldn’t find how/where did the researchers get the concept of “7 mins” and I read several times but I still didn’t find the answer. I like the idea you mentioned why the original article would be more appealing than the one we read and I do agree with the idea you have. People do like to predict things. People would like to get a better outcome in the future. The original article provided significant results but the current study did not indicate any significant results.

  2. Gregory Rosen says:

    At first I was excited to read a replication study, but then wanted to dismiss it given the growing list of limitations. Gradually I came to see how powerful the statistical methods were to keep the ship afloat. This study seems to point toward the fascinating possibility that the children in the famous original Marshmallow Test by Mischel might have resisted marshmallows due to something more nuanced than self-control. I was left to wonder about what Watts et al. meant by the “general and behavioral capacities” that would need to be changed and how intervention developers could do that. The vagueness of that phrase motivated me to search the Internet, where I found a commentary by Barragan-Jason et al.
    Barragan-Jason et al. discuss such capacities as a broader view of patience. They posit that internalization of patience as a social norm during childhood could be a key link to success later in life. The internalization necessitates changing how patient you are according to the situation (Barragan-Jason et al., 2019). In other words, they highlight the importance of the soft skill of social plasticity. I wonder if there is an age cut-off for optimally learning to internalize patience as a social norm, similar to how language learning becomes much more difficult after a certain age.
    Another question is did the study control for exposure to technology, such as digital devices, during childhood. I am biased toward believing that overuse or misuse of digital devices can cause impairment of social skills. Did the learning materials or other items listed under controls for home environment cover this? The home environment controlled for physical environment, and I am curious if this covered natural versus man-made environments. If patience is treated as a social skill, could we extend social context to mean interacting with the greater web of life not limited to other humans? Or is social always limited to humans?
    Overall, this article was a good reminder to be wary of making hasty conclusions about what might enhance a child’s future performance. It also led to a big picture question – does delayed gratification capacity differ according to the cultural factors of cooperation vs competition and sense of abundance vs scarcity?

    • Crystal Quinn says:

      Gregory, I think you brought up really interesting points. The issue of internalization of patience was very intriguing. I could see how exposure to social norms during childhood could be predictive of academic/career success later in life. When I was a social worker, I saw a contrast in social norms in neighborhoods of varying SES. It makes sense to me that children exposed to social norms in which patience is taught and modeled would be able to better manage impulsivity. I would think that it may become more difficult to internalize patience as children age, although perhaps there is no clear cut-off in terms of age. I think it would be interesting to research this issue. I also think your point about the overuse of digital devices is important and definitely can impact social skills as well – I agree that the authors should have considered controlling for this.

    • Chen Tiferet-Dweck says:

      Hi Greg, I agree with your point about technology. The original experiment was done decades ago and today’s technology made everything to be reachable and accessible with minimum effort. I believe that the kids today are considered to be less social (when social skills are measured as face-to-face interaction) because they prefer to use technology to socialize. They prefer texting over talking over the phone, they feel more comfortable to show aggressiveness online etc. Even the way we define social skills today has changed due to technology.
      I think a meta-analysis should compare all Marshmellow studies and examine the effect of technology on these results.

    • drcb says:

      The thought about technology and kids crossed my mind too. Kids may be less used to “doing nothing” these days because they use ipads, etc., (and same for adults too)

  3. Sara Babad says:

    I really appreciated how careful the authors were to outline the limitations of their study and clarify the generalizability of their findings. I did have the following methodological concerns: (1) The authors point out that using a 7-minute maximum time limit for the study led to an unanticipated ceiling effect, but, like Kathy mentioned, they do not explain why they chose the 7-minute ceiling in the first place. Considering that Mischel (1974) used a 15-20-minute cut-point, it is unclear why the authors chose to cut this time by more than half. This is an especially salient question considering that there was a significant difference in which children reached the ceiling based on whether or not their mothers completed college. My second question (2) pertains to Table 1, in which the authors show demographic data for degreed and nondegreed mothers in their sample and for the 1998 ECLS-K national sample. Although they clearly show differences in percentage, I would want to see whether these differences are significant. The authors surmise that the nondegreed mothers are most similar to the ECLS-K sample (with the exception of percentage of participants who are Hispanic), but they could easily have backed this up with chi-square analyses.

    Regarding the article, overall, I was struck by the finding that showed that children who waited less than 20 seconds before eating the marshmallow were significantly different from those who waited more than 20 seconds, and that this second group was no different than those who waited the full 7 minutes. It seems that there is something significant about the 20s second cut-point. It made me think of the “magic number 7” – that people can actively recall +/- 7 items before they start forgetting. Is there a similar “magical” (and I don’t mean actually magical) quality to the “20 seconds” the authors found? The authors conclude that it is far more likely that impulsivity, or self-control, explains ability to delay gratification, more so than the use of cognitive strategies suggested by prior literature. Does this mean that those who wait less than 20 seconds are more prone to impulsivity and those who wait 20 seconds or more are able to overcome their impulsivity? Or are they less impulsive to begin with?
    Furthermore, it did not surprise me that they found that demographic variables like home stability, available resources, and income were significantly different in those who could and could not wait 7 minutes, and more so for non-degreed mothers. To me, this speaks to the impact of low socioeconomic status on delayed gratification in children. I therefore thought that an interesting follow-up study would be to look at adverse childhood experiences (e.g., like growing up with a parent with a mental illness or experiencing any kind of abuse) as another significant childhood stressor and ability to delay gratification. Could these factors, that are known to be related to impulsivity (a facet of executive function) also explain differences in ability to delay gratification? Is there something unique about the kind of adversity that affects delayed gratification that the authors found in this study or are there nuances within adversity?

    • Katherine Chang says:

      I also had a thought on the possibility of a “magic” cutoff point for 20 seconds. At the same time, I was running through some possibilities other than impulsivity (which is definitely could be). I wondered if perhaps the children had trouble understanding the instructions correctly and so they would simply eat the treat. Or what they were tricksters and rule-defiers. Which could also make sense that they might have lower scores on achievement–if they don’t understand (maybe a learning disability) or had problems with authority. While writing this, it feels somewhat outlandish but food for thought.

    • Dakota Egglefield says:

      Hi Sara,

      I also was not surprised by the results you mentioned regarding the impact of low socioeconomic status on delayed gratification in children. Sure, low income or poverty may have an effect on a child’s ability to delay gratification with a task involving food, especially if they are not being fed enough at home. This leads me to think of another variable that should probably be assessed prior to conducting a study like this – state-level satiety or hunger. I can imagine that a very hungry child (low or high SES) would have more trouble delaying gratification with knowledge that they could have a treat immediately rather than wait 7 (or 15, or 20) minutes to have two treats. This is probably more true for the original study because it may prove much harder to wait 20 minutes rather than 7, even for a child who is quite hungry. However, this may yield different results in either condition and would be interesting to include in the covariates.

      • Victoria Fairchild says:

        Hi Sara & Dakota,
        I shared very similar thoughts when reading the article, especially in terms of baseline measures that might help to clarify that the effects of lower SES on delayed gratification. When motivating children with food, especially as they were told they can have their preference between different kinds of snack foods (which they might not be allowed to have or can afford to regularly buy), it might be worth recording time of last meal and baseline reported hunger.
        Additionally, I did think that replicating the study and then choosing to adjust the cut-off from 7 to 20 was a strange choice, especially as it would have been interesting to be able to compare the SES classes and education levels of the parents to timing scores. It seems that the “20 second” cutt-off only really applies to lower SES, as 90% of higher SES children lasted longer than that. I wonder if they had kept the initial timepoint of 20 minutes if a similarly meaningful cutoff might have been established for the higher SES children as well.

    • Scott Ewing says:

      I’d be willing to bet they’d have found some link to adverse childhood experiences, so long as they were thorough in measuring and differentiating ‘experiences’. E.g., those who experienced chronic, physical abuse may have different outcomes re: delaying gratification than those who had to take on the role of caretaker for a loved one.

      I’d be very interested to run these analyses in another 10-15 years, looking at substance abuse (shocker). Addiction hits all SES – it’s not uncommon to find millionaire coke addicts and alcoholics – but treatment adherence/relapse rates are very much correlated with SES. I wonder if the marshmallow test could show a unique contribution to variance in drug use.

  4. Sunghee Kim says:

    As a little child, waiting 7 minutes to get a treat would be very difficult even though a child knows he will get a better outcome (more treats) if he waits. It is also the same for me. If I can patiently study more then, I can get a better result/grade or If I work out 1 more hour then I can have a toned body but I always fail to do so. Being patient/waiting is very difficult even for me although I know that it will help me out to get a better result. I think the article is all about this, more waiting gives you a better outcome; when you are able to do so early in life then you could do in later life.
    I was very happy to see the replication of the very famous study but the current study failed to find out the significant results as the original study did. I personally thought the delay of gratification was too short (7 mins) compared to the original study (more than 10 mins then extended to 20 mins). It was very interesting to find out the results itself. There were no significant results shown. First, my question is why only the researchers interested in mothers but not fathers? What if mother had a college degree but father did not have a college degree? Or what if mother did not have a college degree but father had a college degree? From my perspective, father’s college degree also matters a lot to a child as much as mother’s. Most of the time, mother takes care of a child, that would be the reason why?
    As written, there were many limitations which could manipulate the results. I wish there were more mothers with a college degree in order to make a better comparison with mothers with non-college degree. There were way more non-college degree mothers than mothers with a college degree. I personally did not enjoy reading the article. I did not like the fact that there were so many limitations from the start which were fixable but the researchers did not effortfully fix the issues. I thought it was very an outdated study but it was actually from last year. This article made me read the original study because I was too curious to find out the actual results.

    • minjung park says:

      Hi, Sunghee. First, thank you for sharing your idea. I agree with you. Thinking back to my experience, I also felt hard to wait for minutes to get my favorite snacks. It’s like a disaster. Also, the present study is a longitudinal study, so there might have lots of unexpected causes during the period of study. I completely understand the researchers’ efforts to draw various results. However, as you already said above, there are some limits on this study compared to the original study by Mischel and Shoda. They might know these respects. For future research, they should consider those limitations and better ways to draw valid and reliable results.

    • Shanna Razak says:

      I never understood how a child waiting for a piece of candy shows that they are more likely to have more patience later in life. I believe patience depends on how much you really want the outcome. In other words motivation. For example, me waiting years to obtain a degree is based on my high motivation and ambition further supporting my patience. This is because I know this degree will lead me to become a School Psychologist. However, my patience would decrease if I had to go through years of residency to become a doctor. A career that is very noble and comes with great benefits but I personally wouldn’t have the motivation or patience to pursue.

  5. Sergiu Barcaci says:

    Being in and studying the child psych field led me to initially be very excited upon seeing this as an article. A retest of a popular study from back in the day with improvements and up-to-date controls and limitations? I thought I’d be thrilled and invested while reading this but it was quite honestly a disappointment. The first major issue that wasn’t even given an answer for was the reduction of the previous 15-20 minutes limit to just 7. I’m sure many of us are going to point out the issues with this, so all I’m going to say is my suggestion would’ve been that if they really wanted to leave it at 7 minutes, they could have just made it the soft cap. Meaning they could’ve used the statistics they found for their 7 minute measures while also letting the experiment continue past the 7 minutes just to see if it would’ve had a significant impact or not.
    The next two issues I had were in regard to a specific SES control and the lack of control for degree/non-degree fathers. In today’s age, electronics can be a huge factor in a child’s day-to-day life and they can have a huge impact on executive functions. I’ve browsed for some articles that pertain to that and there aren’t that many, therefore the incorporation of this as a control would’ve also added some more data to the pool. The second is the focus on mother’s being the ones with or without degrees when the father could also be without a degree which would also not only contribute to the SES status but also the way the child is raised and what s/he is exposed to.
    The amount of in-depth measures during the milestones was somewhat reassuring but the present limitations could also skew that data as well, especially when in regard to disruptive, anti-social, rebellious behavior. While they do say that the results of the Applied Problems subtest were practically inconclusive, I do agree that this does posit the potential for this test to not only just be about behavior and self-control, but also as a means to predict cognitive potential. Besides that, the fact that there were no major results is rather disappointing, especially when thinking that some of the things they could have and should have taken into account could have led to some interesting findings, even if it wasn’t what they were initially going for.

    • Bengi says:

      Thanks for the commentary, Sergiu. Your critique about the arbitrary choice of the 7 minute limit is eye-opening for me as I’m not very quick to catch the experimental details. What are these measures of? What is this time representative of? are further questions that make me really curious. I agree with your critique about the other possible factors within the family life and the way the everyday life is being lived that need to be controlled for and I appreciate the emphasis on fathers and digital technology’s diffusion into the lives of children. Maybe more educated parents bring prohibitions and limitations to these times which might tie into this discussion. I didn’t initially think about digital technology but was wondering about the implications of family and immediate social networks (for example what is the influence of having siblings vs. being an only child, having a live-in grandmother or a nanny). From the point of common sense, I know that my cousin became very impatient with her demands from her parents after her sister was born.

    • Katherine Chang says:

      That’s an interesting point that you have about a soft cap. I also had a similar thought that they could have still recorded and seen how long someone could’ve waited. The problem with that is how would you structure the verbal instructions then? The children are prompted that if they waited 7 minutes they would be get more treats and if they cannot wait, they simply need to press a button (or some kind of alert) to signify they want the treat earlier than 7 minutes. I think that might be the limitation to using a soft cap but I would be curious how you might pose the instructions to be able to do this.

    • Gregory Rosen says:

      Hi Sergiu, I was also interested in the effect of digital devices, and I like how you mentioned this in the wider sense of electronics. Here is a link on the cautionary side: On the other hand, electronics could be beneficial to children on the autistic spectrum:

  6. Bengi says:

    The overarching question that is in my mind throughout the article is the idea of correlating gratification delay and academic achievement and noncognitive later outcomes while knowing that there are contextual factors, that are an extension of socioeconomic status and parenting style and environment, that are working as processes that become so connected to one another and embedded in the everyday life that will become difficult to control for, but I appreciate the fact that in the form of “child background” and “home environment” variables were controlled for. While on the one hand, delay gratification might be manifesting as a part of a wider process of parenting systems/styles and separating it might cause missing other variables that are contributing to the enhancement of the later outcomes, other aspect of the problem lies in looking at the outcomes at a “later age” and not something that takes place instantly as in experimental studies; in correlational studies, we are not able to see the direct impact of the variable due to lack of causation and subjects’ being immersed in the everyday life rather than being observed in an isolated form (which is a limitation they already underlined). To add to that, I always think about what else other than how much time children are waiting or when they are giving up, such as possible interactions between the experimenter and child and child’s cognitive and affective processes, might point to a pattern about the way if and to what extent delay takes place. Watts et al (2018) study, by selecting the less educated mother’s children and also through getting a diverse population (which still needed more work with different racial backgrounds), addresses an important drawback in the sample of Mischel and Shoda’s study about being highly selective, but the conceptual basis remains unchallenged. Lastly, reading the “little evidence supporting associations between early delay ability and later outcomes for the higher-SES sample” (1169), made me think about the limitations of thinking about interventions as geared towards delay gratification and cognitive and behavioral capacities without putting them into coordination with wider social policy efforts supporting low-income families.

    • Sergiu says:

      Hey Bengi! There definitely should be a control for how the experimenter interacts with the child. It might be very hard to do but it could eliminate some bias there. Two others I just thought of might also be the gender of the experimenter (a male telling them to wait as opposed to a female), and whether the child ate before doing the experiment (I’m sure a hungry child is more likely to scarf down the marshmallow that’s in front of him). It was interesting that there was little evidence, but could that also be because of one or many of the limitations that were not accounted for?

  7. Dakota Egglefield says:

    Similar to those above me, and at risk of being repetitive, my main point of contention was on the arbitrary 7-minute ceiling and more so the authors’ failure to comment on it aside from it being different from Shoda & Mischel’s original study. I won’t focus on those specific details, but this 7-minute ceiling led me to be slightly confused by the potential directional bias of this study as a whole. Towards the beginning of the article and in the introduction, it seemed like the authors set out with hopeful intention to replicate the marshmallow study to their best ability, additionally with a larger and more diverse sample. Personally, I felt that there was an attitudinal switch once the researchers realized what they were dealing with regarding the results. I am aware that papers are not written in this pre- and post- statistical analyses fashion, but it seemed like upon reporting their results, the authors kept emphasizing that they took every measure they could and ran excess statistical analyses in attempt to soundly replicate the findings, but still there was minimal significance. It frustrated me that they seemed to put so much weight on their efforts to show that their attempt to replicate, but then do not even mention one method that they could have employed to control or correct for the arbitrary 7-minute ceiling. They explain every other aspect of the study in multiple ways, but then just kind of shrug off this ceiling as if there was absolutely nothing they could do about or contribute to this aspect. While they are not responsible for initial study design (this data was taken from a large longitudinal database), they could have even suggested something in their discussion or limitations as simple as what Sergiu recommended, like making 7 minutes a soft cap. I find it hard to believe that the authors could not find one aspect of the arbitrary ceiling to comment on, especially when they were able to so thoroughly comment on all other aspects. It is clear even from our discussion board that this is something most people would notice and want some type of explanation or condolence for.

    An additional point that caught my eye (recency effect from our test, I’m sure) was in the “Alternative outcome measures” section of the results where the authors spoke about their reliance on aggregated measures of achievement and behavior. I immediately thought about Epstein’s theory of aggregation, and his claim that aggregating scores will give you a bigger and better picture of the individual. Mischel’s counterargument to Epstein stated that it did not account for potential meaningful patterns in a person’s behavior, and that we cannot be sure that aggregated scores will apply at the individual level. While aggregation may be a strong tactic for deriving nomothetic research findings, the authors tested separate models for math, reading, and externalizing and internalizing behaviors in order to account for this potential absence of information. There were nearly no statistical differences when looking at aggregated scores versus taking a slightly more idiographic approach and observing the separate subtest scores. What consequences does this small piece of information, nestled into the results, have for the two opposing sides of the personality debate? Speaking to the definition, this is not an idiographic approach, and certainly not what Mischel had in mind when challenging Epstein. However, the fact that there were no different results between the aggregated measures and individual measures in this study serves as evidence for the strength of the technique of aggregation and a nomothetic approach to personality.

    • Sara Babad says:

      Hi Dakota! I appreciate your comment regarding an aggregated measure of achievement and behavior. The child development literature suggests that these two constructs are very nuanced. Considering the significant differences the authors found between children of degreed vs nondegreed mothers, a factor related to socioeconomic status (which is also associated with achievement and behavior), I would be much more interested in a more detailed picture.

  8. Crystal Quinn says:

    I have always found the marshmallow test to be interesting (and cute!). I think that it was innovative to improve upon the original work of Mischel and colleagues, as the authors aimed to better understand the association between delay of gratification and adolescent achievement and behavior. I agree that it is important to consider other contributions to adolescent achievement and behavior, such as mother’s level of education, demographic characteristics (i.e., gender and race), and birth weight.

    I found that it particularly interesting that the steepest incline in achievement score at age 15 was observed after the first 20 seconds, indicating that if a child has immediate impulse control, achievement is dramatically improved at age 15. This finding really highlights impulse control as being important in predicting future academic success, as opposed to solely “self-control”, which is mentioned in the discussion. I think it would be useful to explore what aspects of functioning predict impulsivity in young childhood. I think this information could help clinicians/researchers develop intervention models. I am thinking that the ability to self-soothe could contribute. If this were the case, then perhaps teaching self-soothing skills prior to 54 months could be helpful. I have heard of programs that are offered to new mothers at hospitals (particularly in low-SES neighborhoods), which focus on teaching mothers about effective childcare. I think it would be helpful to add psychoeducation about self-soothing to such a program.

    In terms of criticism, I do not understand why the authors stopped the waiting period at 7 minutes when in the original study by Mischel, children were asked to wait 15-20 minutes. There was substantial evidence that this created a ceiling for the higher SES children. Moreover, it is unclear to me how the authors defined SES. From my brief literature review, I see that 3 components are usually considered when estimating SES: 1) education, 2) income, and 3) occupation. From what I see, it doesn’t appear that all of these components were considered when defining SES. In addition, the article stated that only mother’s income was collected, while I believe that SES is determined by “head of household”, meaning whoever makes a larger salary when 2 parents are involved. I believe this is a limitation in terms of operationally defining SES.

    • Sara Babad says:

      Hi Crystal! Your post made me think about other factors that may be relevant. This study doesn’t touch on this at all, but I wonder if there are cultural differences in the marshmallow test. Would children from collectivist cultures be expected to be better able to refrain from eating the marshmallow?

    • Shira Russell-Giller says:

      Crystal, I am so with you that the author’s defined SES a bit strangely. I didn’t understand why mother’s education level was the main factor, especially since the authors seemed to have access to factors such as income-to-needs ratio, although you make a good point about only the mother’s income being collected. I wonder why the fathers’ income and level of education were not factors at all – it seems like these would be pretty basic factors in understanding SES levels. Since the main grouping variable was mothers’ education level, I wonder if the groups would have been separated entirely differently had the fathers’ education level been taken into account.

  9. minjung park says:

    First of all, I was so happy to make me remind of ‘Marshmallow test’ by Mischel and Shoda that is one of the pronounced articles. The Mischel’s original study gave a field of psychology an impressive and fundamental result about the significant correlation between a child’s ability to delay gratification and socioemotional behaviors. However, I guess the present study has some limitations compared to the original study Mischel and Shoda were conducted. Even though I like the researchers’ effort of the present study to extend and replica the original one with various trials, they couldn’t find the results out significantly.
    First of all, there are some debates about the waiting time, 7 minutes. It is too short to draw the representative result. In the original study, Mischel and Shoda kept the waiting time as 15 minutes to 20 minutes to have more reliable and valid results. Therefore, studying only the observation of 7 minutes might be a lack of precision. Also, I am curious why they only used 7 minutes for observing a child’s ability to delay gratification. Next, This is kinds of longitudinal study that means that they might have multiple causes than just one that affects on results during the long-run study. When it is a study with children as, and the study includes a developmental period from early childhood to adolescence, they might have many things to have effects on child’s development such as parenting, unexpected events, or educational changes, or early intervention of technology. Thus, it is tough and needs many efforts to keep a longitudinal study.
    Through the ‘Marshmallow Test.’ I can say that studying with uncontrolled and normal children, especially estimating children’ ability to wait and control by letting them watch just their favorite snacks for several minutes, is really hard. Thinking back to my early childhood, my mom had to work, and always she put some snacks and notes on the dining table like “Do not eat until you are done with cleaning your room” or “If you finish your school work, you can eat one or two of them.” Honestly, I couldn’t bear it because waiting in front of my favorite dumplings is like a disaster, so I gave rewards to myself first, and then I did homework or clean some stuff. I think my example is kind of different from the present study, but I understand and recognize that making a result with a longitudinal study with uncontrolled children is difficult.

  10. Shira Russell-Giller says:

    Watts, Duncan and Quan’s (2018) revisit of the marshmallow test originally studied by Shoda, Mischel, and Peake (1990) provided new insight into what is known about the intriguing predictions that can be made from early delayed-gratification performance. There are a few points I found quite interesting. For one, I thought that it was interesting that the authors chose to divide the main groups into children of mothers with and without a college education instead of dividing the groups by poverty level. The authors did control for family income and they did point out that for children with non-degreed mothers, the children who were from higher income families completed the delayed gratification task significantly more than the children who were from lower income families, but the main group divider in this study was mother’s education level. I would be interested to know what would have happened if the main groups were divided by poverty, since it seems logical for low income to play a big role in the development of delayed gratification. Particularly when dealing with scarcity, I would imagine that children growing up without easily available basic necessities would not be well-practiced in delayed-gratification, since they would be operating in “survival mode” and grab whatever they can while it is available to them. Sure, mother’s education level is most definitely interlinked with family income, but it is not the same.

    Related to this point, I was thinking about what could possibly account for the fact that for children of non-degreed mothers, children who waited 20 seconds longer than the others had greater achievement in the future. The authors proposed that self-control may be a mechanism involved, although analysis confirmed that it only accounted for some of the delayed gratification effect. Perhaps another mediator between delayed gratification and later achievement outcomes is the degree to which the child is operating from “survival mode.” Children living in an environment where scarcity exists may be more inclined to eat the piece of candy sooner, simply because the availability of the candy now triggers a survival mode response. Most children may debate about whether to eat the candy now or wait so they can have more, regardless of their family’s income level. But perhaps survival mode helps to account for the 20 second wait difference in the delayed-gratification task. The children in survival mode may have caved just a little bit sooner than the rest of the children.

    • Ulzhan Yeshengazina says:

      I also agree with your reaction to this article. It makes complete sense that the level of the social-economic status of the family would play a crucial role in developing skills related to self-control. The study certainly has its limitations, and further investigations required to address these issues.

      I also think that a child’s delay of gratification or impulsivity may be affected by the broken promises that the child experienced in the past in low SES; this may explain (partially) why children may not wait for the second marshmallow.

      Additionally, I would also consider family size. Having a sibling/s may affect a child’s behavior in this context.

      • Sophie Schiff says:

        I completely agree that it would have been interesting to further explore factors such as family size. It seems like the authors had access to more detailed information regarding family life (mean number of children, whether or not the mother was married) as displayed in Table 1, but chose to look at higher level variables to use as controls based on the HOME inventory. The authors indicate that their choice to use the HOME variables as controls is based on prior research that these are highly relevant in cognitive, emotional, and behavioral functioning, but I think a more thorough investigation into more specific factors of home life would have been interesting to include.

      • Ariel Zucker says:

        Shira – I completely agree with your point regarding SES. I was thinking the same thing.

        Along those lines, I wish that that the authors had also taken religion into account. Considering there are some religions that implement some form of delayed gratification before food consumption (i.e., saying grace or other blessings before meals), I would imagine this might provide a potential confound. Specifically, participants who are typically used to saying some form of prayer before eating, might be less impulsive.

      • drcb says:

        Were you thinking that siblings may create greater competition, and thus greater likelihood to quickly eat the treat?

    • drcb says:

      I agree using mom’s education was an unusual delineation. Could it have been used because it created more interesting results?

      Also, I like your survival mode ideas. This could literally relate to issues of hunger in lower SES folks.

Leave a Reply




Spam prevention powered by Akismet

Skip to toolbar