A literature review of articles relevant to my Provost research project.
Linkages Between Self-Regulation and Delinquency The importance of studying juvenile delinquency has grown in the last decade. For example, in 2008, juveniles committed 16% of all violent crimes and 28% of all property crimes in the U.S. (Slowikowski, 2009). According to Slowikowski, juvenile arrest rates for robbery then increased in 2009. These statistics make apparent the urgent need to study and examine juvenile delinquency and its predictors. Many studies have claimed that there is a relation between children and adolescents’ selfregulation and delinquency. For instance, Shekarhar and Gibson (2011) found that adolescents with less self-control were more likely to engage in violent and property offences. Furthermore, researchers have hypothesized that children’s earlier self-regulation is a predictor of later delinquency (King, Fleming, Monahan and Catalano, 2011; Shekarhar and Gibson, 2011). For instance, studies have found that children’s lower self-regulation is a risk factor for later substance use (King et all, 2011). Taking these findings into account, we hypothesize that children’s self-regulation will predict delinquency. In other words, children who have low self-regulation in early childhood may be at a higher risk for delinquent behavior in later adolescence. We will run three types of analyses. First, we will conduct a cross-sectional study of the data to demonstration selfregulation predicting delinquency. Second, we will conduct a lagged study of the data to prove earlier self-regulation predicting later delinquency. Lastly, we will conduct a longitudinal study of the data in which earlier self-regulation predicts later delinquency, while controlling for earlier delinquency.
Measures of Self-Regulation Self-regulation can be assessed in several different ways. There are overall ratings as well as specific measures of self-regulatory skills. A general measure of self-regulation includes the ability to manage one’s own behavior, attention, and emotions (Li-Grining, 2012). More precisely defined, self-regulatory skills involve the ability to delay gratification, restrain inappropriate impulses, focus attention on specific tasks, and participate in problem solving (LiGrining, 2012). This concept of self-regulation has been defined as the child’s ability to stop themselves from a habitual response and change their focus to alternate thoughts or behaviors (Li-Grining, 2012). The term self-control has been used as a global assessment of self-regulatory competence. Self-control has been defined as the ability to inhibit ones impulses (King et al., 2011). In addition to the ability to manage ones impulses, self-control can be defined as attention control, impulsivity, and behavioral undercontrol or disinhibition (King et al., 2011). Poor impulse control has been operationally defined as the “interaction between emotional motivations or the tendency or desire to act on positive or negative urges, and cognitive control, the ability to playfully and actively control those urges and redirect and focus attention to appropriate stimuli” (King et al., 2011). King et al., 2011 stated that multiple studies have found that poor self-control or related constructs such as impulsivity and behavioral disinhibition to be risk factors for drug and alcohol use. More specific ratings of self-regulatory skills include executive function. Executive function is defined as the measurement of higher-order processes that are involved in the selfregulation of learning, emotions, and behavior (Li-Grining, Lennon, Marcus & Haas, 2012). Executive function skills are made up of three different components including working memory,
inhibitory control, and cognitive flexibility (Li-Grining et al., 2012). Working memory is defined as the ability to hold onto information in short-term memory and use it later (Li-Grining et al., 2012). Inhibitory control is the ability to manage temptations (Li-Grining et al., 2012).). Cognitive flexibility refers to the capability to adjust to differences in rules across different contexts (Li-Grining et al., 2012).). Lastly, behavior problems have been used as a proxy for lower self-regulation (Gusdorf, Karreman, van Aken, Dekovc, and Tujil, 2011). Effortful control is defined as an aspect of temperament and consists of delay of gratification, effortful attention and suppression (Gusdorf et al., 2011). Gusdorf et al., 2011 assessed effortful control and its relationship to behavior problems including conduct problems and hyperactivity. Gusdorf et al., 2011 stated “effortful control is the ability to inhibit a dominant response and activate a subdominant response.” Literature Review After having defined self-regulation, we will discuss how self-regulation can be used to explain and predict delinquency. We will cover two types of literature. The first type of literature analyzes middle childhood throughout adolescence. The second type of literature analyzes middle childhood throughout adulthood. Two studies including King et al., 2011 and Shekarhar & Gibson, 2011 fall into the first category of literature- which analyzes children longitudinally from middle childhood to adolescence. Both studies focus on self-control but result in different outcomes. Middle childhood throughout adolescence. King et al., 2011 focuses on predicting substance abuse. King et al., 2011 aimed to examine growth across middle adolescence in two aspects of impulse control, self-control problems and attention problems, and then test the effects of level and change in these variables on levels and change over time in substance abuse. Many
studies have found that poor impulse control is a risk factor for substance abuse. The goal of this research was to test latent growth models of impulse control and connect these variations in impulse control to changes in substance abuse. They hypothesized that higher levels of selfcontrol problems would be associated with higher levels of substance use from middle school to high school and greater increases during this time. Participants were drawn from the Raising Healthy Children project, a longitudinal study focused on the prevention of and intervention on problem behaviors in 10 public Pacific Northwest schools. Data were drawn from an intervention designed to reduce substance use, therefore the effects on the levels and changes of impulse control and substance use were controlled for. Participants for the current study included 955 students with any data from grades 6-11. Eighty-two percent of participants had data from all waves. Fifty-four percent of participants were male. The racial demographics of the sample were 81.6% White, 6.9% Asian/Pacific Islander, 4% Black, and 2.7% Native American. Surveys were administered to the parents, students, and teachers in the spring of each year during the duration of the project. Using self-report measures, the frequency of alcohol, tobacco and drug use were assessed at grades 810. Self-control problems were measured at all five waves, using one self-report measure, and two teacher-report and parent-report measures. Attention problems were assessed using the Teacher Observation of Classroom Adaption-Revised (TOCA-R) and two self-report measures. Results from latent growth models (conditional and unconditional) demonstrate that both the level and the rate of change in self-control and attention problems throughout middle school predicted the level but not the course of substance use at the end of high school. Attention and self-control problems were found to have individual differences over time, suggesting some individuals increase while others decrease. Specifically, results show that participants differed in
the degree of changes in their ability to regulate their behavior. Findings suggest that those individuals with poor impulse control in middle school are at a higher risk for substance use in high school. Adolescents with the lowest reports of impulse control problems in middle school demonstrated more change in substance use as they developed compared to their peers (as teenagers enter the later years of high school, drug and alcohol use escalates). Similar to the King et al., 2011 study, Shekarhar and Gibson 2011 focus on self-controlbut their outcome is concentrated on offending instead of substance abuse. Shekarhar and Gibson (2011) aimed to examine the empirical validity of Gottfredson and Hirschi’s theory of selfcontrol relating specifically to gender and offending behaviors across a large group of Latino youth. Gottfredson and Hirschi’s theory states that differences in self-control are influenced by differences in parenting practices, specifically that parents are more attentive to the behavior of female than male children, which then contributes to children’s levels of self-control. The researchers hypothesize that differences in self-control will help to be accounted for by gender differences- females with more control and males with less. Therefore, males will offend more often than females. Data were drawn from the Project on Human Development in Chicago Neighborhoods (PHDCN), a longitudinal study of Chicago neighborhoods which allows for the study of selfcontrol, parenting, and offending in Latinos. Seven hundred and thirty-nine Latino youth in the 9, 12, and 15 year-old cohorts from waves 1 and 2 were used. Participants were predominantly Mexican (65%), followed by Puerto Rican (20%), and other Latino ethnicities (15%). At wave 1 the average age ranged from 8-16 years, and at wave 2 the average age was 12 years. Fifty-two percent of the participants were males, and 41% of families resided in low socioeconomic areas.
Participants’ offending was measured using a self-report offending instrument in order to assess their violent and property offenses in the past 12 months. Self-control was measured using the Emotionality, Activity, Sociability, and Impulsivity (EASI) scale. The EASI scale was given to primary caregivers to evaluate four temperaments: emotionality, activity, sociability, and impulsivity. In order to evaluate parenting, the Home Observation for Measurement of Environment was used to evaluate three parenting dimensions relative to Gottfredson and Hirschi’s theory: parental warmth, lack of hostility, and supervision. Results from independent t-tests demonstrate that males on average were more likely to be offenders (including violent and property offences) than females. In line with Gottfredson and Hirschi’s theory, males were shown to have less self-control than females. Logistic regression models demonstrate a significant interaction between gender and violent offending- males were significantly more likely to report violent offences than females. Age and being foreign-born were also significantly related- older Latino youth and those who were non-foreign born were more likely to report violent offences. Findings show that parental warmth, hostility, and supervision did not have a significant influence on violent offending. Results show low selfcontrol for both males and females had a positive significant influence on violent offendingthose with less self-control were more likely to engage in violent offenses, and a similar pattern of results was found for property offenses. Early/middle childhood throughout adulthood. The second body of literature in which we will focus on follows children longitudinally from early and middle childhood-adulthood. Three studies will be discussed from that all focus on more specific measures of self-regulation. Two articles focus on executive function and one article focuses on behavior problems.
One article on executive functioning predicting risky behavior was conducted by Pharo, Sim, Graham, Gross, and Hayne (2011). Pharo et al., (2011) aimed to examine the relationship between risky behavior, personality traits, and performance on neuro-psychological batteries of executive functioning in adolescents and young adults. Past research has shown that the risk of injury or death is 2-3 times that of childhood in adolescence. Adolescents are more likely than children and adults to abuse alcohol and drugs, have unprotected sex, engage in antisocial behavior, and drive recklessly or while intoxicated. The goal of this research was to examine relations between real life risk taking behaviors, personality factors, and neuropsychological tests of executive functioning. The researchers hypothesized that differences in executive functioning would independently predict real life risk taking behavior (after taking into account different personality traits). Participants were recruited via word of mouth from a city in New Zealand. Participants for the current study included 136 13-17 year olds (69 males, 67 females), and 57 18-22 year olds (27 males, 30 females). In order to assess personality traits, participants were given the Zuckerman and Kuhlman Personality Questionnaire-Short Form (ZKPQ-F). The ZKPQ-F assesses impulsive sensation seeking, neuroticism-anxiety, aggression-hostility, activity, and sociability. Participants were also given the Life Experiences Questionnaire which consists of three sections. The first section assesses risky behavior, in which participants answer questions relating to different types of risky behaviors. The second section assesses early delinquency and illegal behaviors. The third section assesses alcohol use. Several neuropsychological tests were administered in order to assess executive functioning, including: Controlled Oral Word Association Test, Mental Control and Backward Digit Span scores from the Weschler Memory
Scale-III, Mental Arithmetic Score from the Weschler Adult Intelligence Scale-III, preservative error score from the Wisconsin Card Scoring Test, and the Stroop Color-Word test. Results from ANOVAs revealed that males and adolescents reported significantly higher levels of risky behavior than females and adults. Results from Pearson product-moment correlations between participants’ Z-Scores demonstrate that as participants’ scores on neuropsychological tests increased, their scores on real world risk taking decreased. Furthermore, as levels of impulsivity, sensation seeking, aggression, and sociability increased, so did risky behavior. Results from hierarchical regression analyses revealed that the neuro-function battery was uniquely predictive of real world risk taking once personality traits were taken into account. The second study using executive function as a measure of self-regulation predicting substance use was conducted by Wilens, Martelon, Fried, Petty, Bateman, & Biederman, (2011). Wilens et al. (2011) aimed to assess whether executive functioning deficits were a risk factor for future substance abuse in young adults. The researchers describe executive functioning as a group of higher-order cognitive processes responsible for self-regulatory and goal-oriented behavior. The researchers define a executive functioning deficit (EFD) as having at least 2 abnormal executive functioning neuropsychological tests out of a total of 6. Poor executive functioning has been hypothesized to lead to an inappropriate range of behaviors including the regulation of substance use. Children with neurobehavioral disinhibitions are at an increased risk for substance use disorder. In this study, researchers examined whether EFDs affect the risk for substance use disorder in children with ADHD, as well as children without ADHD. They hypothesize that EFDs identified in childhood would predict substance use disorders in emerging adulthood, with ADHD independent of ADHD status.
Boys and girls were assessed at two different waves between the ages of 6-17 years. Participants were drawn from two longitudinal case-control family studies of children of both genders with and without ADHD, obtained from psychiatric and pediatric resources. Psychiatric assessment relied on the Schedule for Affective Disorder and Schizophrenia for Children (KSADES-E) and the DSM-IV Structured Clinical Interview (SCID). Substance use was assessed using the K-SADES-E and the SCID. Executive functioning was assessed using the Copy Organizations and Delay Organizations of the Rey-Osterrieth Complex Figure, Auditory Continuous Performance Test, Wisconsin Card Sorting Test, Wide Range Achievement of Memory and Learning Test, Stroop Test, and the Freedom from Distractibility Index. In order to measure EFDs, the researchers created a binary measure of executive functioning disorders. They assessed differences at 4 and 5 year follow-ups between those with ADHD and those without using t-tests for continuous outcomes and the Pearson x2 test for binary outcomes. In order to examine the differences between EFDs and substance use the researchers used logistic regressions to test for any developments or new cases of substance use associated with a participant’s current EFD status. The researchers found that those with ADHD were more likely to have EFDs: 41% of those with ADHD did compared to 11% of the controls. Those with ADHD were significantly more likely to be diagnosed with a substance use disorder than controls. Those with ADHD and an EFD were more likely to smoke cigarettes. However, among those with ADHD, having EFD did not increase likeliness of having a substance use disorder. The researchers found no significant differences between those with EFD and controls across all substance use outcomes. Lastly, a study using behavior problems as a proxy for self-regulation to predict later substance abuse was conducted by Holtmann, Buchmann, Esser, Schmidt, Banaschewski, &
Laucht, (2011). Holtmann et al., 2011 aimed to examine the diagnostic and functional trajectories and the predictive utility of the Child Behavior Checklist-Dysregulation Profile (CBCL-DP). Recent longitudinal studies have found that the CBCL-DP in childhood was associated with heightened rates of psychiatric disorders, increased risk for suicidality, and psychosocial impairment. This is the first study to examine the longitudinal course of the CBCLDP, and the researchers expected that the CBCL-DP would represent a wide range of adult psychopathology. The researchers hypothesized that participants with high CBCL-DP scores in childhood would experience higher levels of externalizing/internalizing disorders, increased substance use/abuse, suicidal ideation/behavior, and psychosocial impairment. Participants were drawn from the Manheim Study of Children at Risk, an epidemiological cohort study focused on early risk factors from birth into adulthood. Participants included 325 children of predominantly European descent. One hundred and fifty-one of the participants were males. Children were born between the years of 1986-1988 and assessments were conducted for 19 years at the ages of 2, 4.5, 8, 11, 15, and 19. When children were at age 8, their parents completed the Manheim Parent Interview which reports on psychiatric diagnoses. When children were ages 8 and 11, parents completed the CBCL. When the young adult were at age 19 the Structured Clinical Interview for DSM-IV was administered to assess psychiatric disorders, as well as the Global Assessment of Function Scale to assess overall psychological functioning. In addition, nicotine dependence was measured with the Fagerstorm test for Nicotine Dependence. Lastly, young adults completed the Beck Depression Inventory to assess suicidal ideations. Correlational analyses were used to determine associations of the CBCL-DPwith gender, family income, parental education level, level of functioning and psychiatric diagnosis at age 8.
Results demonstrate that while gender was unrelated to CBCL-DP, significant correlations were found with family income, psychiatric diagnosis, and the child’s level of functioning, and a trend association was found with parental education. Chi square and Mann-Whitney U Tests were used to determine gender differences regarding diagnosis, nicotine dependence, suicidal behaviors and overall psychological functioning at age 19. Gender was found to be unrelated to CBCL-DP results. Almost one third of participants reported daily smoking and 23% met DSM-IV criteria for a diagnosis. Logistic and linear regression models were performed to determine the association between the CBCL-DP and psychiatric outcome in young adulthood. Results demonstrate that at ages 8 and 11 CBCL-DP scores strongly predict several psychiatric outcomes at age 19: the higher the CBCL-DP score in childhood, the more increased risk for ADHD, suicidality, mood and substance use disorders, and poorer overall functioning in young adulthood. Conclusion Limitations and future directions. Despite the strengths of these studies, like all research are there limitations. These limitations suggest directions for future research. Gusdorf et al., 2011 identified that their study demonstrates implications for the assessment of effortful control. Future studies need to be aware of the difference between a more general picture (CBQ) and a more specific picture (ECB) of effortful control (Gusdorf, et al., 2011). Shekarhar and Gibson 2011 stated that future investigations should look more closely at Latinos and the differences between being native and foreign-born, as well as explore differences in Latino families due to the inability of parenting variables in this study being able to explain offending behavior. Wilens et al., 2011 reported future research should examine their findings using their sample to assess a full range of risk and identifying predictors of substance use.
Implications. Holtmann et al., 2011 stated that future studies should focus on and develop intervention strategies in order to change children’s deteriorating development. Holtmann et al., 2011 brings awareness to the importance of timing- intervention in early childhood can make a drastic difference. King et al., 2011 stated interventions in middle childhood should aim to improve the development of self-control during middle school, which may prevent some adolescents from engaging in early substance use. However, earlier intervention may provide broader protective factors (King et al., 2011).