Life-course theory divides the life course into childhood, adulthood, and old age, emphasizes the influence of early factors, believes that the results of human behavior should be traced back to the “upstream” factors of the life course, and the early life course is the most critical impact on the individual, which lasts for a lifetime and is difficult to remedy or eliminate (Bernardia et al., 2019). Childhood adversity (CA) refers to stressful or traumatic experiences that occur during childhood, including physical or emotional abuse during childhood and chronic environmental stress (Dube et al., 2001). A previous study found that childhood adversity is associated with adolescent violence, and individuals who experienced abuse and violence in childhood are more likely to be violent in adolescence and early adulthood (Salo et al., 2021). A specific study using a life-course framework to examine the impact of adverse childhood experiences with internalizing and externalizing problems found that early adverse experiences were significantly associated with problem behaviors (Schroeder et al., 2020). Similar results were found in related studies of Chinese adolescents, where adolescents with more difficult and unpredictable childhood environments had more internalizing and externalizing problem behaviors (Chang and Lu, 2018; Chang et al., 2021). Therefore, we propose:

Life history strategy (LHS) refers to a kind of allocation and trade-off strategy of human beings for various resources. This theory holds that in order to survive and reproduce, individuals will allocate and trade off their own resources when facing limited resource conditions (Gladden et al., 2010). Researchers describe life history strategies as a continuum from “fast” to “slow,” with faster life history strategies associated with earlier age of fertility and sexual development and a preference for immediate satisfaction. Conversely, slower life history strategies are associated with later fertility and sexual development, and a preference for delayed satisfaction (Figueredo et al., 2005; Bielby et al., 2007; Griskevicius et al., 2013). Individuals’ choice of life history strategies is related to early life experiences (Belsky et al., 1991; Kuzawa et al., 2010). Individuals are more inclined to adopt fast life history strategy when their early childhood environment is more difficult and unpredictable (White et al., 2013). Due to the instability of the living environment, individuals are exposed to many risk factors, and using a slow life history strategy will expose individuals to greater risk. Therefore, fast life history strategies are meaningful in harsh and unpredictable ecological environments. In conclusion, the harshness and unpredictability of the environment will motivate individuals to choose a fast life history strategy. Conversely, when the life environment is stable and controllable, individuals prefer a slow life history strategy (Ellis et al., 2009).

Different life history strategy leads to different behavioral outcomes. Individuals who use fast life history strategy pay more attention to the moment, emphasize short-term effects, and are more inclined to take impulsive risk-taking, instant gratification, and even social irregularities; Individuals who use the slow life history strategy pay more attention to long-term development, tend to delay gratification, and obey the law (Figueredo et al., 2004, 2005). Minkov and Beaver (2016) analyzed worldwide indices of criminal violence (homicide, robbery, and assault) and showed that differences in criminal violence in most countries are largely due to differences in life history strategy, and they argue that life history theory is the most compelling explanation for violent crime as a complex phenomenon. Recent research examining the relationship between childhood circumstances, life history strategy, and problem behaviors among Chinese adolescents found that adolescents with more difficult and unpredictable childhood circumstances tended to choose fast life history strategy and tended to have more aggressive behaviors, impulsivity, poor academic performance, internalized and externalized problem behaviors (Chang and Lu, 2018; Chang et al., 2019; Lu and Chang, 2019; Chang et al., 2021). Therefore, we propose:

Dark triad (DT) refers to a group of personality traits consisting of three dark personalities: Machiavellianism, psychopathy, and narcissism, with common manifestations such as aggressive behavior, lack of empathy, and disregard for conventional morality (Paulhus and Williams, 2002). Narcissism is characterized by an exaggerated self-concept, an excessive sense of self, an overwhelming sense of self-worth, and a general lack of concern for others (Vonk et al., 2013). Machiavellianism is characterized by manipulativeness, control, and deception for personal gain (Esperger and Bereczkei, 2012). The main characteristics of psychopathy include apathy, stimulus-seeking behavior, and impulsivity (Brankley and Rule, 2014).

Dark triad may moderate the process by which childhood adversity affects life history strategy. Previous research has found that life history strategy not only influenced by early environment but also by individual personality (Figueredo et al., 2004, 2005). Individuals with higher level of the dark triad are more impulsive, exploitative, focused only on personal interests, more focused on immediate benefits, and desire direct procreative-related benefits (Gladden et al., 2009). Individuals who experience more early adversity gradually develop an implicit attitude regarding the scarcity of resources and the brevity of life (Ellis et al., 2009). Thus, when individuals have difficult childhood environments, individuals with high dark triad choose fast life history strategy to obtain more resources (Jonason et al., 2014; Csatho and Birkas, 2018; Geng et al., 2021); whereas individuals with low dark triad have an advantage in delayed gratification and are good at careful planning (Jonason et al., 2010), which helps them to choose a slow life history strategy. Therefore, we propose:

In addition, dark triad may moderate the process of life history strategy on problem behaviors. Individuals who adopt fast life history strategies tend to pursue immediate benefits (Figueredo et al., 2004). The dark triad represents the dark side of personality, and individuals with high levels of the dark triad are more cynical, immoral, manipulative, and will do anything to achieve their goals, even if they do not conform to social norms (Jonason et al., 2015, 2017). Influenced by the high dark triad, individuals tend to do anything in their pursuit of immediate benefits and are more likely to have problem behaviors; therefore, individuals with the high dark triad will have more problematic behaviors when using fast life history strategies (Jones and Paulhus, 2010; Paulhus et al., 2022). In contrast, individuals with a low dark triad are more focused on the long term and are good at careful planning, which can effectively compensate for the lack of long-term consideration in the fast life history strategy (Paulhus and Williams, 2002). Therefore, individuals with a fast life history strategy show fewer problem behaviors under the influence of the low dark triad. One study found that the dark triad moderated the effects of fast life history strategy on aggressive behavior, and individuals who used fast life history strategies had significantly more aggressive behavior when they had higher levels of the dark triad (Shi, 2021). Therefore, we propose:

To summarize, this study examined the combined effects of childhood adversity, life history strategy, and the dark triad on the problem behaviors (Figure 1), using new street corner youth on campus as participants and constructing a moderated mediation model to explore the mediating role of life history strategy and the moderating role of the dark triad in the relationship between childhood adversity and problem behaviors.

To ensure that the results of the study were representative, data were collected from multiple schools. Considering the differences in the management methods of public schools and private schools, when selecting schools for research, they are also divided according to the nature of the schools. Among them, there are 8 junior high schools, including two private middle schools and six public middle schools; there are 4 senior middle schools, including one private high school and three public high schools; there are three secondary vocational schools, including two public secondary vocational schools and one private secondary vocational school; there are three colleges and universities, including one public school, two private schools, and one private undergraduate college. High schools develop knowledge-based students, while secondary vocational colleges develop application-based students who can go to work directly after graduation and have higher admission scores in high schools than in secondary vocational colleges.

The experimenter, as a psychology teacher, surveyed 19 schools over an 8-month period from April 2021 to December 2021 through semi-participant observation, interviews, and fieldwork. During this 8-month period, we mainly observed and interviewed on-campus students who interacted with off-campus street corner youth. During the survey, the experimenter purposefully selected new street corner youth on campus to complete the questionnaire. Our selection criteria for new street corner youth on campus were students who had interpersonal interactions with street corner youth, such as being friends with street corner youth, partying, etc. Before students answered the questionnaire, the experimenter would explain the precautions and read out the instructions. A total of 657 questionnaires were distributed, and 637 valid questionnaires were collected, with a valid return rate of 99.96%. The average age of the participants was 15.05 years old (SD = 0.94), with 429 males and 208 females. Data were collected in an anonymous manner after reading the instructions and signing the informed consent form, participants completed paper-version questionnaires within 15 min. After completing the survey, everyone was received a small gift. The authors’ university ethics committee approved the current research.

Childhood adversity was measured using the Childhood Environment Scale developed by Ellis et al. (2009), and revised by Chen and Chang (2016). The scale was validated into Chinese language (Zhang and Liu, 2022) and is a 14-item questionnaire consisting of two dimensions (hardship and unpredictability), such as, “How many times in your childhood did you seek medical attention for injuries from beatings?” Participants rated all items on a 4-point scale (1 = never to 4 = always). With higher scores indicating greater hardship and unpredictability of childhood adversity. In this study, the Cronbach’s ɑ was 0.97.

The life history strategy was measured using the Life History Strategies Scale developed by Fiueredo et al. (2006), and revised by Wang et al. (2017). Although the original version of the questionnaire has received some criticism in recent years (Gruijters and Fleuren, 2018; Manson et al., 2020), the Chinese version of the scale revised by Geng et al. is widely used in China and has good reliability (Zhang and Geng, 2021). The scale contains 20 items scored and is a uni-dimensional scale on a 7-point scale (1 = Completely disagree to 7 = Strongly agree), such as, “I avoid taking risks.” With a high score indicating a slow life history strategy. In this study, the Cronbach’s ɑ was 0.97.

The dark triad was measured using the Dark Triad Scale developed by Jones and Paulhus (2014), and revised by Geng et al. (2015). The Chinese version of the scale revised by Geng et al. is widely used in China and has good reliability (Yang et al., 2021). It is a 12-item questionnaire consisting of three dimensions (Machiavellianism, Psychopathic, and Narcissism). Participants rated all items on a 7-point scale (1 = Totally disagree to 7 = Strongly agree), such as, “I do not care much about whether my actions are ethical or not.” With higher scores indicating higher predisposition to the dark triad personality. In this study, the Cronbach’s ɑ was 0.92.

Descriptive statistics and correlations of key variables were analyzed using SPSS 26.0. The analysis of moderated mediation model was performed using PROCESS macro (Model 59). Prior to the analyses, the predictors were standardized. In all analyses, we included gender and age as a covariate.

We addressed the possibility of common method variance, the results revealed that there are 10 factors with eigenvalues greater than 1, and the explained variance of the first factor was 24.68%, which is lower than the specified maximum of 40%. Therefore, common method variance was not a significant concern in this study.

The results of the descriptive statistics and correlation analysis are shown in Table 1. The scores of PB for new street corner youth on campus in this study were higher than the scores of adolescents in previous studies (M = 0.59, SD = 0.25), indicating that there were more problem behaviors for adolescents in this study than for the general youth (Liu et al., 2020). The scores of CA for the new street corner youth on campus in this study were lower than the scores of adolescents in previous studies (M = 17.97, SD = 0.34), indicating adolescents in this study did not have more difficult and unpredictable childhood circumstances compared to general youth (Sai et al., 2020). The scores of DT for the new street corner youth on campus in this study were higher than the scores of adolescents in previous studies (M = 20.51, SD = 12.08), indicating that the dark triad personality tendencies of adolescents in this study were higher than those of general adolescents (Gao et al., 2020). The scores of LHS for the new street corner youth on campus in this study were lower than the scores of adolescents in previous studies (M = 114.81, SD = 16.17), indicating that adolescents in this study adopt faster life history strategy compared to general youth (Sai et al., 2020).

The correlation analysis between the variables revealed that CA was negatively associated with LHS and positively associated with PB, these results are consistent with previous studies (White et al., 2013; Salo et al., 2021). LHS were positively correlated with PB, this result is inconsistent with previous study (Figueredo et al., 2004). DT was positively correlated with PB, this result is consistent with previous study (Li et al., 2016). CA was positively associated with DT, this result is consistent with previous study (Zhan et al., 2021).

First, model 4 of the SPSS macro program Process was used to test the mediating effect of LHS between CA and PB while controlling for gender and age. The results show (see Table 2 for details) that CA can significantly positively predict PB, CA can significantly negatively predict LHS; after incorporating LHS into the regression equation, CA can still significantly positively predict PB, and the direct effects of CA on PB and the relationship between LHS. The Bootstrap 95% confidence interval of the mediation effect does not contain zero, indicating that the partial mediating effect of life history strategy between CA and PB is significant (ab = −0.05, SE = 0.01, 95%CI = [−0.08, −0.02]). According to the research method of Wen and Ye (2014), when ab and c’ moved in opposing directions, it indicated that LHS had “suppressing effects” on the relationship between CA and PB, and the effect size is |ab/c’| = 18.39%. Overall, effect sizes are modest.

A Moderate Mediation Model of the SPSS macroprogram Process was used to test the moderating effect of DT, controlling for gender and age, and the results were found (see Table 3): the interaction term between CA and DT significantly and negatively predicted LHS; the interaction term between LHS and DT significantly and negatively predicted PB. Thus, DT moderated the first half as well as the second half of the pathway of LHS mediating CA and PB.

Simple slope tests were performed by taking the values of DT plus and minus 1 standard deviation. Figure 2 shows that the negative predictive effect of CA on LHS was significant when the level of DT was low (β = −0.15, t = −2.41, p < 0.05), and the negative predictive effect of CA on LHS remained significant when the level of DT was high, but the increase was relatively slowed (β = −0.29, t = −7.03, p < 0.001). Figure 3 shows that the positive predictive effect of LHS on PB was significant when levels of DT were low (β = 0.28, t = 6.43, p < 0.001).