Research Summaries: Can Adolescents Learn Self-control? Delay of Gratification in the Development of Control over Risk Taking
Today’s research summary is based on a paper by Angela Duckworth and colleagues, and examines the nature of self-control as assessed by risk-taking, sensation-seeking, future time perspective and delay of gratification in US adolescents.
- Adolescents are known to indulge in risk taking activities like recreational drug use and various theories abound as to why adolescence is a particularly sensitive time.
- As per one theory, there is a dopamine surge in reward centers of the brain during adolescence which leads to impulsive sensation seeking behavior. Traditionally, it is believed that the prefrontal cortex , which can override such impulsive behavior, does not mature in teenage and continues to mature till late thirties, and thus unable to self-regulate behavior in the teenage adequately.
- The above view posits that there is not much one can do about impulsive and risk taking behavior as the brain will take its own sweet time to mature; another view suggests that there are two independent processes involved in risk taking behavior- an underlying propensity to indulge in impulsive sensation seeking behavior (which can be considered as the accelerator moving one towards risk taking behavior) and an ability to delay gratification in service of long term goals (which can be considered as the brakes which moves one away from risk taking behaviors).
- Literature review suggests that sensation seeking is uncorrelated with delay of gratification and both may independently impact risk taking behavior. It was unclear from prior research if delay of gratification can be an effective brake even in adolescents who were very high on sensation seeking. Also future time perspective, or the tendency to think about future more than present, is related to reduced risk taking, but the effect may be mediated by ability to delay gratification (because that ability directly depends on an ability to visualize the future) .
- 900 US adolescents were administered a delay discounting task (choice between larger reward later and a smaller reward now) to ascertain their ability to delay gratification. Their sensation seeking and future time perspective was measured using self-report measures. Risk taking was again measured using self report about three risky behaviors viz cigarette smoking, marijuana use, and binge drinking.Structural equation modeling was used to determine the relation between all variables.
- As expected, sensation seeking in teens and delay of gratification were uncorrelated; delay of gratification predicted less risk taking behavior, future time perspective also predicted less risk taking behavior , but not over and beyond its impact on delay of gratification. Sensation seeking peaked around age 18 and then started decreasing; future time perspective kept increasing with age; and temporal discounting showed an upward trend with age.
- For teens that were high in sensation seeking, their temporal discounting increased with age more sharply. The authors explained this due to the fact that teens who were high in sensation seeking would indulge in more risky behavior and on getting negative feedback from environment on these behaviors will learn to self-regulate and increase delay of gratification.
- From this research it seems there are at least two routes to increase your temporal discounting muscle and hence reduce your risk taking behavior. The first approach is to become explicitly future focused and have a stronger future time perspective; the second approach is to explore, experiment and learn from your mistakes as your risk taking backfires. If done in a conducive environment, like graded driving tests, then this can lead to good outcomes.
I found the paper pretty interesting as it clearly dissociates the tow mechanisms that lead to risky behavior. If you found the above interesting, check out the paper here.
- Temporal discounting or time preference is the preference people show towards immediate short-term rewards over higher but later long-term gains. People are willing to accept much lower sums (of say money) now, than they would, for sure, receive at some time in the future. This preference is for sure sums and is distinct and different form uncertainty/risk avoidance.
- Different people have different temporal discounting rates; some discount future gains much more steeply than others – these people will prefer immediate rewards much more strongly than those who have a less steep discount function.
- Typical rewards considered in temporal discounting studies are monetary rewards; however a case can be made that other non-equivalent types of rewards exist like edible items, vacation experiences, health outcomes etc. Previous research has shown that contrary to classical economics models, people have different discount rates for different types of rewards; this is called domain-specificity of temporal discounting.
- Different people desire and like different types of rewards to different degrees; for e.g., someone may desire to be healthy and prioritize over monetary rewards. Although, as per research done by Berridge et al, liking and desire are different functions, they are treated together in this paper and operationalized as temptation for the reward.
- There is evidence that there are two systems involved in decision-making – the system I or ‘hot’ and system II or ‘cold’ popularized by Kahneman et al. The beta-delta preference model formalizes this by positing that there are two factors influencing choice- a beta factor making a sharp distinction in present and future and a constant delta discount factor.
- If you like and desire a reward very much, your emotional/ ‘hot’ system will get activated and will override the ‘cold’ system to the extent that you will discount this reward very steeply (prefer strongly the immediate reward) . If however, you are not too excited by the reward and are indifferent to it, the ‘cold’ system will be much more dominant and discounting will not be as steep.
- The experiment conducted of three reward conditions- eating candy, eating chips, drinking beer and temptation was measured using self-report for these rewards.
- Temporal discounting was measured using a choice task in which choices were presented for different quantities of all three rewards (plus dollars) and the delay contrasted with now, versus a delayed reward at time ranges form one week to 3 years.
- What they found were that were indeed subgroups of people like chip lovers (those who were tempted more strongly by chips than say beer) who also discounted chips more strongly; similarly their discount rates was steeper for chips only and not so for beer.
- Thus, they conclude that discount rate depends on how tempting you find that reward and there is no one domain independent discount rate. In other words, temporal discounting is domain specific. What is discounted steeply by a chip lover (guess, guess, its chips!) is not discounted that steeply by beer lover and vice versa.
- This is important imho as it shows that if you want to counter a particular temptation or distraction, you have to be cognizant of that domain and work within that domain.
If you find papers like these, that are at intersection of economics and psychology interesting do check out the full version that is present online.
This research summary is similar to the earlier one where self-control predicted overweight status; Angela and team have co-authored a similar paper, though based on a different data set and controlling for more confounds.
- Self-control is a variable of concern as ” In this obesogenic context, self control, the capacity to regulate behavior, attention, and emotion in the service of personal standards and goals, is required to forego immediate gratification and choose instead options that protect against weight gain.”
- Weight control may be important for teens, not only for its long term health associations, but also because of its impact on physical attractiveness.
- This study was a prospective longitudinal study that looked at over 100 children in a school setting, and measured their self control and BMI while in grade 5 (mean age 10.5) and correlated it with their BMI when in grade 8.
- Self-control was measured using a variety of methods. Students filled 2 self-report measures of self-control: The Impulsivity subscale of the Eysenck I6 Junior Questionnaire and The Brief Self-Control Scale. Parents as well as teachers also filled the informant version of Brief self-control scale. Apart from this Kirby Delay-Discounting Rate Monetary Choice Questionnaire was used to present hypothetical choices between small reward now and large reward later, meant to judge the delay of gratification. Also an actual behavioral delay of gratification task was used to ascertain self-control. A composite measure was created from these measures.
- Potential confounds like demographics (SES), Happiness (measured by SSLS and PANAS-C) and Intelligence (Otis- Lennon School Ability Test—Seventh Edition Level F) were measured and controlled for in the analysis.
- The authors replicated their earlier result that low self-control in childhood, indeed leads to weight gain in transition to adolescence. High self-control, on the other hand, protects children form weight gain.
Research Summaries: Self-Control Protects Against Overweight Status in the Transition from Childhood to Adolescence
Grit and self-control are the two character strengths on which Angela Duckworth focuses a lot, and this research summary is about a paper co-authored by Angela that shows how a lack of self-control can lead to obesity and weight gain in adolescence; while being more self-controlled helps one stay leaner.
- The authors define self-control as ” the ability to override impulses in order to achieve goals and maintain standards”. It is also the ability to resist short term temptations and distractions in service of long term benefits.
- Self-Control has many positive associations like increased life expectancy, higher report card grades and achievement test scores and career success. However not many have looked at whether and how self-control may be related to the right amount of body weight.
- Weight gain from childhood to adolescence is natural, but excessive weight gain that leads to high BMI (body mass Index) is problematic and associated with negative outcomes like coronary diseases, diabetes etc. some risk/ protective factors like Socio economic status (SES) and pubertal development are well established; however not much work has been done linking personality variables like self-control with excessive weight gain.
- As children enter adolescence they start exercising more and more autonomy regarding their lifestyle choices like when to eat , sleep etc. Self-control, or the ability to delay short term gratification in view of long term well-being, thus becomes a salient feature for them with regards to how they manage their weight.
- This study was a prospective longitudinal study that looked at nearly 850 children in a birth cohort, and measured their self control at age 9 and correlated it with their BMI at age 15.
- Self-control at age 9 was operationalised using informant ratings by mother, father and teacher on the items related to self-control on the Social Skills Rating System (SSRS)questionnaire. Overweight status was established by classifying those with BMI z-scores falling above 85th percentile as overweight.
- Other potential confounds like intelligence, pubertal status etc were also measured and used in the analysis.
- The results showed that the overweight children (at age 15) were half a standard deviation lower on self-control (at age 9) that the normal weight children. Similarly, those children who showed higher self control than average at age 9, were less likely to become overweight at age 15.
- The study is important because it points to one mutable, and under one’s control, factor that leads to excessive weight gain – Self control. Thus, this factor , self-control, can and should be taught during the childhood to adolescence transition. It will not only help the obesity epidemic but will lead to other gains too!
If you want to dig deeper, here is the original article.
This research summary will be especially attractive to those who have interest in psychometric and would like to see how the concept and measure of grit has evolved. In this paper, Angela Duckworth refines her measurement of grit and establishes the test-retest stability of the concept apart form predictive and consensual validity.
- The authors wanted to come up with a briefer version of the grit scale, which would have better internal consistency and still retain the predictive power and the two factor structure of Consistency of Interests and Perseverance of effort.
- As such they dropped 2 items each from both the factor items and retained just 4 items each for each of the factors. The decision to drop the items was taken on the basis of analyzing data across four studies as delineated in their earlier paper, whose research summary is present here. The newer 8 item Grit scale called Grrit -S retained predictive power and showed the same two factor structure.
- Similar to their earlier paper, they did an online study measuring grit-S/grit -O, big five traits as predictors and career changes and educational levels as outcome variables. Grit -S correlated with conscientiousness, but was still able to predict the outcomes over and beyond conscientiousness.
- Using the same online procedure, they asked subjects as well as informants (their friends family members) to complete the Grit-S/Grit -O measures and established the consensual validity of the scale.
- In another study with students, they measured Grit-S/Grit -O for two consecutive springs and established the test-retest stability of the scale as well as its predictive validity where GPA obtained was an outcome variable and so was the number of hours watching television.
- The next study was similar to the West point study they had done for earlier paper, but with grit-S predicting who makes it through the beast barracks.
- The last study was again a followup study of the national spelling bee competitors, this time with a new cohort, and using a new scale and led to similar results, whereby girt predicted who reached which round etc. based partly on who practiced how much and had prior experience participating.
- So, if you were looking for some more areas/ examples of the predictive power of grit, this doesn’t add much to what Anagela et al had presented in the earlier paper, but it does reconfirm finding with a shorter measure that also appears to be a better measure of grit.
so, if you are the one who is fascinated by how scales evolve, do check out this paper here.