Which of the following is a critical element that underlies the development of self-control?

How does effortful control (and related self-regulatory traits) develop from childhood through adolescence? One way to understand development is to evaluate mean-level changes, that is, changes in the average trait level of a population. Previous research suggests that, on average, children’s effortful control tends to increase in early childhood, from age 2 to 6 (Carlson, 2005, Eisenberg et al., 2010, Kochanska and Knaack, 2003, Kochanska et al., 2000, Li-Grining, 2007) and during the transition into adolescence from age 6 to 12 (King et al., 2013, Murphy et al., 1999). Historically, the patterns of mean-level change in self-regulatory traits across adolescence (i.e., age 13–18) have been somewhat mixed. Although it was initially thought that self-regulatory traits would continue to increase from late childhood to young adulthood (Branje et al., 2007, Donnellan et al., 2007, Roberts et al., 2006), more recent research has found either no changes in mean-level effortful control during adolescence from age 12 to 16 (Klimstra, Hale, Raaijmakers, Branje, & Meeus, 2009), or even decreases over time from late childhood to late adolescence, ages 10–17 (Borghuis et al., 2017, De Fruyt et al., 2006, Laceulle et al., 2012, Leon-Carrion et al., 2004, Soto et al., 2011, Van den Akker et al., 2014). In line with the latter findings, previous work with the dataset used in the present study also found decreases in effortful control from age 10 to 16 (Atherton et al., 2019, Atherton et al., in press). Considering these recent findings, the current consensus among many researchers is that a “self-regulatory dip” occurs from late childhood to late adolescence (De Fruyt et al., 2006, Klimstra et al., 2009, Soto and Tackett, 2015, Soto et al., 2011, Tackman et al., 2017, Van den Akker et al., 2014). From late adolescence to young adulthood and beyond, people are expected to resume the well-established, mean-level increases in self-regulatory traits that are consistent with the maturation principle (Bleidorn et al., 2009, Blonigen et al., 2008, Damian et al., 2018, Helson and Moane, 1987, Murphy et al., 1999, Roberts et al., 2003, Roberts et al., 2006, Robins et al., 2001, Van den Akker et al., 2014). The temporary “self-regulatory dip” during adolescence is presumably due to the profound biological, psychological, and social changes that occur during this period (Soto & Tackett, 2015).

Despite the somewhat mixed findings regarding mean-level change patterns in self-regulatory traits across adolescence, which may be partly attributed to differences in how different components of self-regulation change over time (Atherton, Lawson, & Robins, under review), one consistent finding across the literature is the extensive within-person variability in change patterns, whereby some people increase, others decrease, and others remain stable in their self-regulatory trait levels. Thus, even if mean-levels showed decreases in effortful control over a certain developmental period, it is possible to test for correlates of positive change at the individual level.

Indeed, an individual-difference approach is necessary because mean-level changes, or “normative trajectories”, cannot explain why some children successfully develop the capacity to self-regulate, whereas others continually give in to their impulses and fail to persevere toward their goals. Which factors contribute to these individual differences? One possibility is that youths need resources and practice to develop their self-regulatory abilities (Denissen, van Aken, Penke, & Wood, 2013), and opportunities for practice can be influenced by the youths’ social environment and the demands placed on them. Some of the largest demands placed on developing youth are in the school environment. In fact, the school environment is akin, in many ways, to self-regulation interventions, given that children are required to enact self-regulatory skills to comply with social and structural rules within the classroom, complete homework in a timely manner, etc. Previous research has shown that effortful control co-develops with school-related experiences such as homework and school behavioral problems (Atherton et al., 2019, Atherton et al., in press, Goellner et al., 2017, Ludtke et al., 2011, Tackman et al., 2017), but no research has investigated the co-development of effortful control with outside-of-school demands, such as household chores. The lack of relevant research is surprising given the widespread belief among parents (and even politicians, such as the Spanish parliament) that chores foster responsibility and other self-regulatory traits in their children. So how might the demands of chores contribute to the development of effortful control across childhood and adolescence?

The TESSERA framework (Wrzus & Roberts, 2017) posits that long-term personality development results from repeated short-term situational processes (TESSERA sequences) coupled with associative and reflective processes. Short-term TESSERA sequences consist of Triggering situations, Expectancies, States/State Expressions, and Reactions. Triggering situations are any events or daily situations that are external to the person and place an environmental pressure to respond. Parents assigning household chores is a potential triggering situation, because the assignment usually comes with a set of expectations for appropriate behavior. Expectancies represent the known set of appropriate or useful behaviors associated with a triggering situation (either self- or other-imposed) that serve to achieve a goal. Household chores come with such expectancies because there are specific performance standards associated with each chore that can be clearly explained by parents or observed by the child. States and State Expressions are momentary thoughts, feelings, and behaviors that occur in the specific situation, after expectancies have been triggered. For example, when it comes to chores, children might engage in deep thought on best ways achieve the task goals or they might not; regarding feelings, they might feel enthusiastic about doing a good job or they might feel bored; regarding behaviors, they might persist and aim for good performance or they might give up. Reactions follow state expressions, and they can come either from oneself (reinforcement through emotions) or others (reinforcement and punishment, e.g., verbal feedback, facial behavior such as smiling or frowning, or monetary incentives) in response to states. For example, after completing their chores, youth may feel pride/shame in their performance (i.e., for matching/mismatching their state/expressions or behaviors with the task expectancies) or may be praised/admonished by their parents. With this, the short-term TESSERA sequence ends. A crucial element in transforming such short-term processes into long-term personality development is repetition. Regarding chores, repetition of the short-term TESSERA sequences is likely, because parents are likely to assign chores more than once (e.g., on a weekly basis), which may be one reason as to why chores may influence self-regulation development. Repeatedly experiencing TESSERA sequences can lead to long-term personality change through associative or reflective processes. Associative processes imply that over the course of many repetitions, triggers become associated with behaviors while the self-concept is also activated, and thus, over time, changes in implicit self-concept may occur, thus leading to personality change. Reflective processes presumably change and maintain personality by consciously thinking about one’s past experiences, behavior, thoughts, and feelings. Thus, regarding chores, it is possible that after experiencing the TESSERA sequence multiple times (e.g., a scenario where task persistence and good performance at undesirable chore activities led to pride or praise), children may start integrating “persistence in the face of competing tasks” into their self-concept, which may eventually lead to lasting changes in personality traits such as effortful control, where regulating impulses and behaviors are key components. Notably, the above sequence of processes does not happen in a vacuum. Instead, person-environment transactions (Buss, 1987) play a key role, because people select and create different environments in correspondence with their personality, which means that it is due to personality traits in the first place that personality (in)congruent triggering situations and states can occur. For example, children who are higher in effortful control to begin with might be given or ask for more chores from their parents, which would then set in motion the processes described above and result in increasing engagement with chores across time, as well as increasing levels of effortful control. Thus, it is possible that the co-development of chores and effortful control is bidirectional. In this paper, we tested the co-development of chores and effortful control from age 10 to 16.

The second key goal of this paper was to test to what extent trajectories of change in effortful control (from late childhood through adolescence) would predict work outcomes in young adulthood, above and beyond initial levels of effortful control. We are not aware of any research linking effortful control in childhood/adolescence with later work outcomes, but we review below research on related self-regulatory traits.

As mentioned earlier, most previous research has used statically measured self-regulatory traits (i.e., trait levels measured at one time point) as prospective predictors of work outcomes. Thus, multiple longitudinal studies have found that adolescents who are higher on self-regulatory traits tend to have more successful job outcomes later in life. For example, constraint (i.e., the tendency to endorse social norms, act in a cautious and restrained manner, and avoid thrills) measured in late adolescence prospectively predicts higher occupational attainment, higher job satisfaction, higher work commitment, better job fit (i.e., how well an employee is suited for their position), more material benefits, and more financial independence in early adulthood (Le et al., 2014, Roberts et al., 2003). Higher levels of self-regulatory traits, such as constraint and mature personality measured in adolescence, have also been found to predict higher occupational prestige in young adulthood (Damian et al., 2015, Roberts et al., 2003).

But, why would higher levels of self-regulatory traits measured in adolescence predict better work outcomes in young adulthood? Several different processes have been put forward to explain this relation. For example, the link between self-regulatory traits and later work outcomes may reflect ‘‘recruitment effects,’’ whereby people are selected into jobs/education based on their self-regulation levels. In fact, most employers prefer people who are more organized, responsible, punctual, and industrious (Cook et al., 2000, Sackett and Walmsley, 2014). Even early levels of self-regulatory traits in childhood and adolescence are likely essential because recruitment effects start emerging early in development. For example, children’s personality traits begin to influence their relationships with teachers, which snowball into later educational attainment and work outcomes (Birch & Ladd, 1998). Self-regulatory traits may also impact job fit and job satisfaction, and lower job stress, because people who are more conscientious take a more active role in shaping their work environment to better fit their personalities (Bell and Staw, 1989, Lepine et al., 1997, Roberts et al., 2006, Wrzesniewski and Dutton, 2001). In turn, increased fit with work environments predicts better performance and lower job stress (Harms, Roberts, & Winter, 2006). Another mechanism through which self-regulatory traits may predict work outcomes, is “attrition” or “deselection pressures,” whereby people leave (or are released from) jobs that require self-regulation (e.g., those that require high levels of autonomy, attention to detail, punctuality) (Cairns and Cairns, 1994, Caspi et al., 1998, Kokko et al., 2003, Pulkkinen and Kokko, 2000). Finally, self-regulatory traits may predict work outcomes, and especially job performance, because they consistently predict stronger goal setting and self-efficacy (Erez and Judge, 2001, Judge and Ilies, 2002).

A notable disconnect in the literature is that theory suggests dynamic processes governed by person-environment transactions, whereas empirical research has focused so far largely on static links between self-regulation measured at one point and work outcomes measured at a later point. But what about self-regulation changes across time and their impact on outcomes? We know that people’s self-regulation levels are malleable and change across time, and many intervention programs aimed at children and teens hope to boost self-regulation to improve future work outcomes. However, apart from only two studies (Allemand et al., 2018, Converse et al., 2018), there is no empirical research investigating the role of self-regulation change trajectories in predicting later work outcomes over and above initial levels of self-regulation. Of these two studies, the first investigated the role of self-control development from age 16 to 29 on work outcomes at age 29, and work outcomes included job satisfaction and income (Converse et al., 2018). The second study investigated the role of self-control development from age 12 to 16 on work outcomes at age 35, and the work outcomes included occupational self-efficacy beliefs, achievement motivation, and willingness to further training and development (Allemand et al., 2018). Both studies found evidence for an incremental role of self-control change (over and above initial levels) in predicting work outcomes, such that increases in self-control predicted better work outcomes. However, these previous studies also had some drawbacks. Specifically, both studies used self-reports only to assess self-control, and they called for future studies to use a multi-method approach, where self-regulatory traits are assessed using both self- and informant-reports to boost their validity. Moreover, intelligence was also self-reported in one of the studies (Allemand et al., 2018), parental socio-economic status was missing from the other study (Converse et al., 2018), and the types of work outcomes assessed across the two studies were limited to job satisfaction and work engagement. Finally, the age range covered only included adolescence, but not late childhood, which is an important omission because self-regulation intervention programs supported by policymakers focus on childhood intervention (Cavadel et al., 2017, Heckman, 2012, Murray et al., 2016). We addressed these gaps and extended the findings in the present study.

We used data from a 10-year longitudinal study of Mexican-origin youth (N = 674) who were assessed at ages 10, 12, 14, 16, and 19, with retention rates of 86–92%. The goals of the study were twofold. First, we tested whether household chores and effortful control co-developed from age 10 to 16. Second, we tested whether initial levels of effortful control (age 10), as well as increasing trajectories from age 10 to 16, predicted better work outcomes in young adulthood (age 19).

Regarding the co-development of chores and effortful control, we hypothesized that higher initial levels of chores would be associated with higher initial levels of effortful control, higher initial levels of chores would be associated with steeper increases in effortful control, higher initial levels of effortful control would be associate with steeper increases in chores, and steeper increases in chores would be associated with steeper increases in effortful control from age 10 to 16. Moreover, because girls tend to do more chores than boys (Duckett, Raffaelli, & Richards, 1989) and tend to show higher effortful control levels (Atherton et al., 2019, Atherton et al., in press, Else-Quest et al., 2006), we tested whether gender moderated the co-development of chores and effortful control. Additionally, given our unique sample of Mexican-origin youth, we were able to test another potential moderator of individual differences in co-development: nativity status (i.e., born in the U.S. vs. born in Mexico). We did not have specific predictions regarding these moderation analyses, and thus, they are exploratory. It is possible that cultural norms associated with nativity status may impact the associations between chores and effortful control, because youth born in Mexico (who may be less acculturated) experience both higher expectancies from their parents to perform chores, as well as more satisfaction from performing chores if they see it as a duty to their family. These differences may result in developmental processes that are more intense for those where social norms are pushing them more towards those behaviors and thus, the co-developmental pathways between chores and effortful control may be stronger for youth born in Mexico.

Regarding the prospective impact of effortful control development on work outcomes and based on previous work on related self-regulation constructs, we hypothesized that higher initial levels and steeper increases in effortful control would be associated with more positive work outcomes (i.e., employment, job satisfaction, job fit, job autonomy, satisfaction with job performance, and lower job stress and work-relationship conflict). We also tested whether the link between effortful control development and work outcomes varied by gender and nativity status. Although we did not have clear predictions regarding moderation, we thought exploratory analyses could inform future research, as one could argue that early boosts in effortful control might benefit people’s work outcomes differently based on their demographic characteristics (e.g., if women are expected to exhibit higher levels of effortful control in the work place, then perhaps a boost would be more beneficial for them; or conversely, because men tend to start lower in effortful control as children, perhaps a boost would be more beneficial for them).

This study makes several novel contributions. First, this is the first study to test the co-development of chores and effortful control. Second, this is one of only three studies (the others being Allemand et al., 2018, Converse et al., 2018) to test the prospective association between self-regulatory change (above and beyond initial levels) and work outcomes; in this context, this is the first study to use self-regulation change trajectories assessed from late childhood through adolescence, and it is the first study to focus on effortful control, which is a particularly relevant trait in this age group. Third, unlike the previous two studies on the topic (Allemand et al., 2018, Converse et al., 2018), which only used self-reports, we used both self- and parent-reports of effortful control to reduce the influence of method variance. Fourth, we statistically controlled for family socio-economic status (SES) and child intelligence (measured at age 10) in all analyses; including these covariates was essential because previous research has shown that parental SES and intelligence impact work outcomes, and self-regulatory trait development (e.g., Damian et al., 2015, Shanahan et al., 2014). Fifth, we measured a wide variety of work outcomes in young adulthood (age 19), including employment status, job satisfaction, job stress, job fit, job autonomy, satisfaction with job performance, and work-relationship conflict (of these outcomes, only job satisfaction was assessed previously in relation to self-control change). Having multiple outcomes available in the same data set is important for comparing the magnitude and robustness of the associations across different types of job outcomes. Sixth, this is the first study to test moderating effects of gender and nativity status, which have been previously theorized to be associated with effortful control (Else-Quest et al., 2006, Garcia-Coll and Marks, 2011). Seventh, this is the first study to use a low socio-economic status (SES), ethnic minority sample, where people tend to join the workforce early. Specifically, we used a sample of Mexican-origin youth, which is important for several reasons: (a) diversity in psychological samples is essential for ecological validity; (b) Mexican-origin youth are an understudied but rapidly growing ethnic minority group (U.S. Census Bureau, 2010); and (c) as the labor force becomes increasingly Hispanic/Latino, there is a dire need to better understand the developmental precursors in occupational outcomes in this population.