Abstract
This study explored predictors of adolescent problem behavior utilizing panel data from the National Youth Policy Institute. Data were collected from June 2013 to August 2013 from 3195 middle school students living in 15 cities of South Korea. Data were analysed using hierarchical regression in two steps. The first step included gender and school record; the second step added emotional difficulty, parent-child relationship, peer relationships, self-esteem, materially oriented attitude, and appearance-oriented attitude. Results showed that male students had greater problem behaviors than female students. Greater emotional difficulties were associated with greater problem behaviors. On the other hand, a better parent-child relationship and better peer relationships were related to fewer problem behaviors. Notably, greater materially and appearance-oriented attitudes were related to greater problem behaviors. This study implies that excessively extrinsic values or attitudes should be discouraged to prevent teenager problem behavior.
Keywords
Adolescent problem behavior includes various areas such as delinquency, school bullying, disappearance from home, smoking, drinking, and immoral or illegal activities. According to Problem-Behavior Theory (Jessor & Jessor, 1977), problem behavior means illegal and unsocial behavior. Importantly, problem behavior is closely related to teenager’s mental health and may influence their early adult development.
Problem-Behavior Theory proposes that adolescent problem behavior results from interactions of personal factors (personality system) and environmental factors (perceived environment system) (Donovan & Jessor, 1985). The personality system is an individual’s social cognitive characteristics such as personal values, attitudes, beliefs, and orientations towards self, others, and the world. In particular, this theory suggests that some variables such as academic motivation, self-esteem, and autonomy directly influence teenager problem behavior. The perceived environment system is one’s personal perception of one’s social context. This theory also especially insists that parental support or control, and peer support or approval, directly impact adolescent problem behavior.
Parent-child relationships appear to directly influence teenager problem behavior (Chung, Oh, Moon, & Cho, 2005; Lee, Han, Park, & Roh, 2015). According to numerous studies, excessively authoritarian and protective attitudes of parents tend to negatively influence parent-child relationships, and consequently are more likely to trigger adolescent problem behavior (Rey & Plapp, 1990). These parenting attitudes may cause difficulties in communication between parent and child, and may negatively impact the acquisition of autonomy in the teenager (Pettit, Laird, Dodge, Bates, & Criss, 2001). Furthermore, adolescents who fail to become autonomous are likely to show problem behavior to compensate for this lack of autonomy.
Numerous studies have reported that peer relationships are associated with teenager problem behavior (Fleming, Catalano, Haggerty, & Abbott, 2010; Laird, Jordan, Dodge, Pettit, & Bates, 2001). Conflicts with peers may increase the level of tension in school life, and consequently can cause problem behavior. Furthermore, teenagers who either have no friends or feel they are being treated unfairly by friends are unlikely to sufficiently internalize conventional rules and regulations, and consequently, are prone to showing problem behavior.
Academic record should also be considered as predictor of adolescent problem behavior (Spriqqs, Lannotti, Nansel, & Haynie, 2007). Poor academic achievement is a significant stressor among teenagers, and can negatively impact adolescents’ self-perceptions and self-esteem, which consequently can influence problem behavior (Gu, Lai, & Ye, 2011). In particular, South Korean adolescents experience high levels of stress due to competition for university entrance during middle and high school.
The influence of self-esteem on teenager problem behavior is unclear. Some studies have reported that lower self-esteem is related to more problem behaviors in teenagers (Rosenberg & Rosenberg, 1978), while other studies have suggested that there is no relationship between self-esteem and problem behaviors of youth (Lee, Bell, & Watson, 2007; Wells & Rankin, 1983).
Emotional difficulties are also more likely to be associated with adolescent problem behavior (Craighead, 1991; Puig-Antich, 1982). In particular, depression appears to increase the likelihood of problem behavior among adolescents. In one study, approximately 20% to 30% of juvenile delinquents also appeared to have depression (Chiles, Miller, & Cox, 1980). In addition, high level of anxiety and anger among youth may trigger problem behaviors (Park & Kim, 2014; Rapee & Melveille, 1997).
Problem-Behavior theory emphasizes the importance of personal values and attitudes in understanding adolescent problem behavior. In the same vein, adolescents’ specific internalized values may positively or negatively influence their behavior. On the other hand, few studies have identified the direct impact of teenager values and attitudes on problem behavior. We live in a time when extrinsic values (e.g. materially oriented attitudes) are emphasized. Additionally, because teenagers lack critical thinking ability and psychological maturity compared to adults, they are prone to internalizing extrinsic values from family and surrounding individuals.
Materially oriented attitude (materialism) means the belief that possession of money and goods is the most important factor in happiness and success in one’s life (Belk, 1985; Richins, 1990). Importantly, adolescents with greater materially oriented attitudes will pay less attention to other important areas of their lives (e.g. self-development, interpersonal relationships), and their life satisfaction may decrease. Consequently, they may be more likely to manifest problem behaviors (Csikszentimihalyi & Rochberg-Halton, 1981).
The influence of a materially oriented attitude on adolescent problem behavior has been identified in Western studies, and results have indicated that a greater materially oriented attitude is associated with more problem behaviors in teenagers (Auerbach et al., 2010; Willians, Cox, Hedberg, & Deci, 2000). On the other hand, little research has identified the relationship between materially oriented attitudes and problem behaviors among Asian adolescents.
One of the values that influence adolescents’ lives is their attitude towards appearance. From the time that students enter middle school, they become increasingly more concerned about their appearance, which they spend a lot of time trying to improve. Furthermore, teenagers who believe that appearance is a very important factor in happiness and success, and who therefore over-emphasize its value, will experience high levels of stress and daily tension related to their appearance, and consequently are more likely to manifest problem behaviors (Agnew, 1992). However, little research has focused on this potentially important issue.
Finally, gender should be considered to further understand problem behaviors of teenagers. In previous studies, male students were more likely to manifest problem behaviors such as committing delinquent, illegal activities than female students (Leadbeater, Kuperminc, Blatt, & Hertzog, 1999).
The purpose of this study was to explore important predictors of adolescent problem behavior, and specifically to identify whether materially oriented attitudes and appearance-oriented attitudes independently influence teenager problem behavior.
Method
Participants
Descriptive statistics (N = 3195).
Measures
All of the measures analysed in this study are inventories developed by the NYPI. In order to verify the construct validity of the measures, each one was subjected to exploratory factor analysis. Details of each measure and the results of the associated factor analysis are explained in the respective sections below.
Problem behavior
Teenager problem behavior was assessed by an inventory that consists of ten items. Each item is scored on a five-point scale (always, mostly, frequently, sometimes, never). Higher scores are related to more problem behavior. Examples of items are as follows: ‘I raise money in ways that are not right, to have fun with my friends’, ‘I have done a bad thing I saw in a movie or on TV’. Based on the KMO measure (0.885) and Bartlett’s test of sphericity (p < 0.001), the data are appropriate for factor analysis. Considering the eigenvalues and scree plot, we determined that a two-factor structure was proper, and named the first factor immoral and illegal action and the second factor delinquent. The explanatory variance of the first factor was 36.70%, that of the second factor was 5.56%, and the total explanatory variance was 42.26%. Cronbach’s alpha for the current sample was 0.822 (first 0.830, second 0.642).
Emotional difficulty
Emotional difficulty was measured by an inventory that consists of ten items. Each item is scored on a five-point scale (always, mostly, frequently, sometimes, never). Higher scores are related to more emotional problems. Examples of items are as follows: ‘I often feel sad and gloomy for no reason’, ‘I recently felt anxious for no reason’, ‘I feel angry with myself or people around for no reason’. Based on the KMO measure (0.921) and Bartlett’s test of sphericity (p < 0.001), the data are appropriate for factor analysis. Considering the eigenvalues and scree plot, we determined that a two-factor structure is appropriate, and named the first factor depression and anxiety and the second factor anger. The explanatory variance of the first factor was 45.46%, that of the second factor was 4.55%, and the total explanatory variance was 50.51%. Cronbach’s alpha for the current sample was 0.881 (first 0.887, second 0.695).
Self-esteem
Self-esteem of teenagers was measured by an inventory that consists six items using a five-point scale (always, mostly, frequently, sometimes, never). Higher scores are related to higher self-esteem. Examples of items are as follows: ‘I think of myself as a very capable person’, ‘I think that I am worthwhile’, ‘I think that I can solve my problems on my own’. Based on the KMO measure (0.794) and Bartlett’s test of sphericity (p < 0.001), the data were appropriate for factor analysis. A two-factor structure was identified. We named the first factor positive self-perception, and the second factor negative self-perception. The explanatory variance of the first factor was 25.60% and that of the second factor 40.60%, and the total explanatory variance was 66.21%. Cronbach’s alpha for the current sample was 0.814 (first 0.872, second 0.767).
Parent-child relationship
Perceived parent-child relationships were assessed by an inventory that consists of ten items using a five-point scale (always, mostly, frequently, sometimes, never). This inventory consists of two sub-factors (positive parent-child relationships and negative parent-child relationships). Positive parent-child relationships measure support and acceptance of parents and intimacy/communication between parent and child (e.g. ‘my parents and I understand each other well’), and consists of five items. Negative parent-child relationships measure rejection and indifference of parents and lack of communication between parent and child (e.g. ‘I get stressed because I disagree with my parents’), and consists of five items. Higher scores mean better parent-child relationships. Based on the KMO measure (0.854) and Bartlett’s test of sphericity (p < 0.001), the data were appropriate for factor analysis. The explanatory variance of the first factor was 42.73%, that of the second factor was 11.75%, and the total explanatory variance was 54.48%. Cronbach’s alpha for the current sample was 0.871 (first factor 0.870, second factor 0.835).
Peer relationships
Peer relationships were measured by an inventory that consists of ten items using a five-point scale (always, mostly, frequently, sometimes, never). This inventory is composed of two sub-factors (positive peer relationships and conflicted peer relationships). Positive peer relationships measure friends’ and peers’ support, acceptance, and intimacy (e.g., ‘friends understand me well’). Negative peer relationships measure rejection and unconcern of friends and conflicts with peers (e.g. ‘I have received a bitter curse from a friend’). Higher scores mean better peer relationships. Based on the KMO measure (0.811) and Bartlett’s test of sphericity (p < 0.001), the data were appropriate for factor analysis. The explanatory variance of the first factor was 28.87%, that of the second factor was 16.80%, and the total explanatory variance was 45.67%. Cronbach’s alpha for the current sample was 0.871 (first factor 0.769, second factor 0.802).
Materially oriented attitude
Materially oriented attitude was measured by an inventory that consists of five items using a five-point scale (always, mostly, frequently, sometimes, never). Higher scores are related to a higher materially oriented attitude (‘I think that money is very important in society’, ‘I think entering good schools and having a good job are a means to earn easy money’). Based on the KMO measure (0.687) and Bartlett’s test of sphericity (p < 0.001), the data were appropriate for factor analysis. A one factor structure was identified and the explanatory variance was 44.53%. Cronbach’s alpha for the current sample was 0.675.
Appearance-oriented attitude
Appearance-oriented attitude was measured by an inventory that consists of ten items using a five-point scale (always, mostly, frequently, sometimes, never). Higher scores mean higher appearance-oriented attitude (‘I envy people who have an excellent appearance’, ‘I would be happier if my appearance was better than it is now’, ‘appearance is an important element of success’). Based on the KMO measure (0.927) and Bartlett’s test of sphericity (p < 0.001), the data were appropriate for factor analysis. A one factor structure was identified and the explanatory variance was 50.19%. Cronbach’s alpha for the current sample was 0.907.
Analysis
We conducted a hierarchical regression analysis to identify the influence of the independent variables on adolescent problem behavior using the SPSS 18.0 program (Table 3). In the first step, we included gender and school record, and in the second step, we added parent-child relationship, peer relationships, emotional difficulty, self-esteem, materially oriented attitude, and appearance-oriented attitude.
Results
Results of hierarchical regression
Gender and school record were entered in the first step of hierarchical regression analysis, and the total variance explained was significant at 1.4% [F (3,3062) =21.33, p < .001]. The remaining variables were inserted in the second step, and the incremental variance explained was significant at 30.1% [F (7,3055) = 176.34, p < 0.001]. The total explanatory variance in the final model was 31.5%.
Correlation among variables.
Note: 1 = problem behavior; 2 = emotional difficulty; 3 = parent-child relationship; 4 = peer relationships; 5 = self-esteem; 6 = materially-oriented attitude; 7 = appearance-oriented attitude.
p < 0.05; **p < 0.05, ***p < 0.001.
Results of hierarchical regression analysis.
Note: R2 = adjusted R2; β = standardized coefficient.
p < 0.05; **p < 0.05; ***p < .001.
Emotional difficulty was the strongest independent predictor of teenager behavior problems. More emotional difficulty (higher levels of depression, anxiety, and anger) was closely related to more problem behavior (more delinquent, immoral, and illegal activities). Male students showed more problem behavior than female students. On the other hand, adolescents who have received more support and acceptance from parents showed lower problem behavior. Additionally, teenagers who have experienced more support and acceptance from peers showed lower problem behavior. Finally, teenagers who think money and appearance are very important factors in life showed more problem behavior.
Discussion
The aim of this study was to identify predictors of adolescent problem behavior. Specifically, we explored whether materially oriented and appearance-oriented attitudes independently impact teenager problem behavior.
Consistent with previous studies, male students showed distinctly greater problem behaviors than female students (Carroll, Houghton, Wood, Perkins, & Bower, 2007). School record was significant in the first step of the hierarchical regression analysis, but became non-significant in the second step when personal and environmental variables were added. This result implies that school records may be related to adolescent problem behavior, but that the impact is likely small compared to that of other important factors.
Numerous studies have reported that adolescents’ emotional difficulties such as depression, anxiety, and anger are associated with more problem behavior (Galanaki, Polychronopoulou, & Babalies, 2008; Lee, & Bae, 2015). In this study, emotional difficulty was the most powerful predictor of adolescent problem behavior.
As shown in previous studies, positively perceived parent-child relationships were related to fewer adolescent problem behaviors (Lee, Wesbecher, Lee, & Lee, 2015). As emphasized by social control theory (Hirschi, 2002), if communication between parent and child is lacking, parents have difficultly properly supervising their children, and as a result, the likelihood of teenager problem behavior becomes greater.
Similar to the findings of previous studies, higher satisfaction with peer relationships was associated with fewer adolescent problem behaviors. Based on general strain theory (Agnew, 2012), adolescents who complain of conflicts with peers, and who perceive unfair treatment by friends are likely to experience high levels of tension, and consequently are more likely to manifest problem behavior. This result implies that peer support and good relationships with friends can help to prevent problem behavior.
Similar to findings from recent national cohort studies, self-esteem did not have a significant impact on problem behaviors of youth (Bae, Lee, & Lee, 2015; Lee & Bae, 2015). Almost all previous studies that explored the influence of self-esteem on problem behaviors employed small samples in restricted regions. Therefore, the results of this study, which was conducted with a national sample, imply that the impact of self-esteem on teenager problem behavior may be small compared to that of other key variables (Lee, Bell, & Watson, 2007). In addition, self-esteem might be a mediator variable and indirectly influence problem behavior through other factors rather than being an independent variable directly influencing teenager problem behavior (Swenson & Prelow, 2005).
One remarkable result of this study was that materially oriented attitude was a significant predictor of adolescent problem behavior. This implies that teenagers who believe that money and goods determine happiness and success are more likely to engage in immoral and illegal activities (Kasser, 2002). Because teenagers who focus on the pursuit of money or material goods are likely to take advantage of other people to acquire them, an excessively materialistic attitude could cause conflicts with peers and problem behavior.
Previous studies of Western individuals have indicated that teenagers’ materialistic attitudes can cause problem behaviors (Pirog & Roberts, 2007). However, little research has been conducted with Eastern individuals (Auerbach et al., 2010). Asian nations such as South Korea, China, and Japan have experienced rapid economic development in a relatively short period. Thus, individuals from these nations may highly appraise the importance in life of money and material goods, and teenagers of these nations may uncritically internalize materialistic attitudes. Thus, verifying the influence of materially oriented attitudes on the problem behavior of Asian adolescents appears important to enhance understanding of the field.
The finding that a greater appearance-oriented attitude is related to greater problem behaviors is of interest. Adolescents with a greater appearance-orientation are likely to suffer from stress related to their appearance, and are more likely to lack interest in various important life domains (e.g. relationships, conventional values, and self-development). Thus, this attitude may cause self-dissatisfaction and conflicts with other people, and consequently may trigger adolescent problem behavior.
Self-determination theory proposes that three crucial needs (relationship, competence, and autonomy) have an important role in our well-being (Ryan & Deci, 2000). Adolescents who focus excessively on extrinsic values (materialism/appearance-orientation) can easily ignore these crucial needs, and are much likely to experience dissatisfaction with life in the long term. Consequently, they are more likely to show emotional or behavioral problems. Additionally, they may be likely to pay less attention to academic achievement and could leave school without completing their education.
These results imply that maintaining good relationships with parents and peers and the emotional support of surrounding individuals may be important primarily to prevent adolescent problem behavior. Another important issue raised by this study is that teenagers who strongly believe that money and appearance determine happiness and success in life may be more likely to devalue other important life domains such as interpersonal relationships and self-fulfillment, and may be less likely to internalize important social or intrinsic values (e.g. academic motivation, self-understanding). Furthermore, they may experience conflicts with friends and people around them more easily, and consequently, be more likely to commit immoral or illegal actions. Thus, parents and schools need to educate adolescents in the importance of intrinsic values such as self-development and self-understanding rather than extrinsic values such as materially oriented and appearance-oriented attitudes.
The results of this study may have implications for researchers in settings outside of South Korea. That is, researchers from Western countries (e.g. France, Australia, and America) may need to take into account culturally relevant values as well as personal and environmental factors such as emotional difficulties and quality of relationships to deeply understand adolescent problem behavior. For example, freedom of individual expression is highly valued in Western countries. Thus, researchers may hypothesize that an excessively appearance-focused attitude or extreme individualism may influence adolescent problem behavior in these countries.
Further research is needed to identify the specific causal pathways among the relevant variables (i.e., emotional difficulty, parent-child relationship, peer relationships, materially or appearance-oriented attitudes, and problem behaviors), because teenagers’ values and attitudes are affected by parents and peers, and emotional difficulties of youth are closely related to relationships with his parents and friends. Finally, a limitation of this study is that the sample included only middle school students, so caution should be used when generalizing from the results.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
