Abstract
Parents influence urban youths’ violence-related behaviors. To provide effective guidance, parents should understand how youth perceive conflict, yet little empirical research has been conducted regarding parent and youth perceptions of conflict. The aims of this article are to (a) report on the nature of discrepancies in attribution of fault, (b) present qualitative data about the varying rationales for fault attribution, and (c) use quantitative data to identify correlates of discrepancy including report of attitudes toward violence, parental communication, and parents’ messages about retaliatory violence. Interviews were conducted with 101 parent/adolescent dyads. The study population consisted of African American female caretakers (n = 92; that is, mothers, grandmothers, aunts) and fathers (n = 9) and their early adolescents (mean age = 13.6). A total of 53 dyads were discrepant in identifying instigators in one or both videos. When discrepancy was present, the parent was more likely to identify the actor who reacted to the situation as at fault. In the logistic regression models, parental attitudes about retaliatory violence were a significant correlate of discrepancy, such that as parent attitudes supporting retaliatory violence increased, the odds of discrepancy decreased. The results suggest that parents and adolescents do not always view conflict situations similarly, which may inhibit effective parent–child communication, parental advice, and discipline. Individuals developing and implementing family-based violence prevention interventions need to be cognizant of the complexity of fault attribution and design strategies to promote conversations around attribution of fault and effective conflict management.
Introduction
Violence is a serious threat to the health and safety of young people in the United States, particularly racial and ethnic minorities (Sampson, Morenoff, & Raudenbush, 2005). Data from a nationally representative survey of youth in the United States found that 32% of students reported they were in a physical fight one or more times in the 12 months preceding the survey (Centers for Disease Control and Prevention, 2010). In 2009, homicide was the second leading cause of death among all youth aged 10 to 18 years, and the leading cause of death for Black youth (Centers for Disease Control and Prevention, 2009). The rate of violent crime is highest in urban areas (National Criminal Justice Referral Service, 2013). Therefore, identifying effective interventions to help reduce or prevent youth involvement in violence, particularly for minority youth, is critical.
Youth violence is a complex public health problem with a growing body of research that highlights the need for prevention strategies across multiple levels of influences (Office of the Surgeon General, 2001). This article will broadly focus on the role of parents in preventing urban youths’ aggression and violence-related behaviors (Copeland-Linder et al., 2007; Orpinas, Murray, & Kelder, 1999). As primary socializing agents for their children, parents reinforce appropriate attitudes and behaviors regarding violence in the home and at school (Roche, Ahmed, & Blum, 2008; Wright & Fitzpatrick, 2006). In this role, parents provide “coaching” on how to handle actual or potential violent situations (Kliewer et al., 2006; Lindstrom Johnson, Finigan, Bradshaw, Haynie, & Cheng, 2013). Yet little empirical research has been conducted to understand whether parents and youth perceive conflict situations in the same manner. This article intends to inform this critical gap in understanding of parent and youth perceptions of conflict situations, which has implications for improving parental communication, advice, and discipline around conflict.
Parent Influences on Youth Involvement in Violence
According to the social cognitive theory (SCT), parents play a key role, through observational learning and incentive motivation, in the development of youths’ biased social information processing and subsequently their likelihood of attributing hostile intents to others in ambiguous social interactions (Bandura, 1977, 1986). Parental attitudes toward violence significantly influence youths’ attitudes toward violence as well as their involvement in violence. In one study, Kliewer and colleagues (2006) reported that among the African American youth in their sample, youth who engaged in aggressive coping, defined as using aggression to deal with violent situations, most often had parents who both suggested and modeled aggressive coping. In a separate study of parents and their 12-to-17-year-olds presenting at an emergency department, parents’ attitudes predicted youths’ aggressive behavior even after controlling for youths’ attitudes (Solomon, Bradshaw, Wright, & Cheng, 2008). Similarly, a school-based study of middle-school students found that children’s perceptions of parents’ attitudes toward fighting were the strongest predictor of aggression (Orpinas et al., 1999).
Informant Discrepancies
In recent years, researchers have begun to examine the potential implications of parent–child discrepancies in reports of youth and parent perceptions or behavior. The existing literature generally focuses on parent and child reports of child psychopathology (Athay, Riemer, & Bickman, 2012; Mascendaro, Herman, & Webster-Stratton, 2012) and parenting practices (i.e., family communication, parental monitoring; De Los Reyes, Goodman, Kliewer, & Reid-Quinones, 2010; Guion, Mrug, & Windle, 2009). For example, among early adolescents, discrepancies in parent–youth reports of parental monitoring were associated with greater levels of internalizing behaviors (i.e., anxious, withdrawn depression; Han et al., 2012) and greater levels of problem behavior (i.e., delinquency, substance use; Lippold, Greenberg, & Feinberg, 2011). Discrepancy will continue to be a growing area of interest because differences in parent–child reports are not only considered a reflection of the quality of parent–child interactions but they also may suggest potential for poor child outcomes (e.g., risky driving behavior, increased internalizing symptoms, deficits in social competence; Guion et al., 2009). The discrepancy literature has not yet focused on proximal processes, including the interpretation of social situations and what disagreement may mean for youth behaviors. However, one study that has examined these processes was conducted by Kliewer et al. (2006). The authors recruited 101 female caregiver–child dyads and showed them a film clip depicting community violence. The dyads were then asked to respond to several questions including how the parents of the boys in the clip might provide them coping assistance and what the boys in the video should do to cope. Coaching, caregivers’ modeling of their own coping, and family context were significant influences on youths’ suggested coping strategies.
Current Study
Urban parents living in violent neighborhoods may provide inconsistent messages about conflict avoidance to their children; a previous analysis of data from this study suggested parents both promoted non-violent conflict resolution and described situations in which their children would need to use violence (Lindstrom Johnson et al., 2013). We extend those findings in the present study, by examining discrepancies in how parents and early adolescents interpret situational peer violence conflict situations. Past research has not explored the similarities between parent and early adolescents’ interpretations of conflict situations. This is an important area of exploration because the effectiveness of parents’ messages about violence is predicated on parents and youth interpreting conflict situations similarly. Discrepancy may be problematic because parents’ messages about fighting and conflict avoidance may not correspond with youths’ violence-related experiences. However, parents and youth identifying the same instigator is not as important as how the dyads are talking about conflict and understanding the determinants of blame (i.e., social information processing; Crick & Dodge, 1994). In this article, we will (a) report on the nature of discrepancies in attribution of fault, (b) present qualitative data about the varying rationales for fault attribution, and (c) use quantitative data to identify correlates of discrepancy including report of attitudes toward violence, parental communication, and parents’ messages about retaliatory violence.
Method
Participants
The data for this study were collected between August 2007 and November 2008. Eligible participants had a child enrolled in one of three urban public middle schools that had previously worked with the research team on a school-based group mentoring randomized trial. The participating middle schools served a significant number of low-income students (i.e., > 70% qualified for free/reduced meals), were on probation for persistently dangerous status defined by the No Child Left Behind criteria of high suspension rates for behavior problems, and were in neighborhoods characterized by high levels of unemployment and violent crime. Also, eligible participants had lived in their neighborhood for at least 6 months, were English speaking, and were reachable by telephone. Parent–youth dyads who both agreed to participate were included in the study. Of the 307 families contacted, 144 completed data collection, 137 families had full-data available for secondary analysis, and 101 families were included in this analysis. Thirty-six transcripts were excluded because one or both participants did not clearly identify an instigator: participant(s) provided an unusable response (26) or the transcript did not include participant’s response (i.e., inaudible response or responses from one participant; 10).
Procedure
The majority of the data collection occurred in the participants’ homes, with a small number of parents requesting data collection take place in a community location. The study was approved by the Institutional Review Boards of the Johns Hopkins University, the Eunice Kennedy Shriver National Institute of Child Health and Human Development, and the city school district research office. Participants provided written consent or assent prior to participation.
During data collection, the parent and youth each separately completed an audio-facilitated questionnaire, which allowed them to privately listen to questions and record their answers. The questionnaire asked parents and youth about demographic characteristics, youth involvement in risk activities, and attitudes about violence.
After completing the questionnaire, the parent and youth were brought together for a semi-structured interview to discuss brief video clips depicting teens in common conflict situations. The video clips were taken from a video series created by Strategies Against Violence Everywhere, a nonprofit organization focused on preventing youth violence (SAVE, 2010). In Video 1, an adolescent girl (locker dweller) is standing at her locker in a crowded hallway when another girl walking by bumps into her (passerby), making the locker dweller drop her schoolbooks. The locker dweller then shouts, “Why you knock my books down? Can’t ya’ll at least say sorry?” after which the passerby turns around and pushes her. A fight quickly ensues. In Video 2, a group of girls is walking outside. They are talking, laughing, and dancing, when one girl trips and falls to the ground. The other girls start laughing, which upsets the girl who fell because she does not think it is funny that she almost got hurt. One of the “friends” (taunter) tells her it is funny that she fell down. In response, the girl who fell picks up a rock and throws it (rock thrower) at the taunter. A third video depicting a conflict between boys on a basketball court was also viewed but was not included in these analyses because there was almost complete agreement as to the instigator. The video portrayed a violation of good sportsmanship, which may represent a type of conflict situation that poses very little ambiguity regarding who began the conflict. In this sample, parents and youth perceived a clear instigator.
Every dyad watched each clip together and through a semi-structured interview, they were asked to discuss what they saw in the video. The questions included the following: “Describe what you saw happen in the video,” “What happened that started the fight?” “Who started the fight?” and “Why do you say that?” Interviews were balanced with regard to whether the parent or child was prompted to start the conversation. Conversations were audio taped and then transcribed by a professional transcription service.
Quantitative Measures
Youth reported their sex, age, and race/ethnicity, while parent reported their sex, age, race/ethnicity, and education level.
Youth and parent attitudes toward retaliatory violence were measured using a scale developed by Hill and Noblin (1991). Youths’ attitudes were assessed using eight items (youth Cronbach’s α = 0.76); sample items included, “If someone hits you, hit them back” and “It is okay to hurt people if they hurt you first.” Parents’ attitudes were assessed using six items that paralleled those asked of youth (parent α = 0.72). Youth and parents responded using a 4-point Likert-type scale measuring agreement. Some items were reverse coded and the items were summed such that a higher score indicated more aggressive attitudes.
Family communication was assessed by both youth and parents using a 10-item scale developed by Barnes and Olson (youth α = 0.89; parent α = 0.85; Barnes & Olson, 1985). Sample items from the youth survey included, “My parent is a good listener” or “My parent expresses affection to me.” Parents’ attitudes were assessed using items that paralleled those asked of the youth. Youth and parents responded on a 5-point Likert-type scale, and responses were summed such that a higher score indicated better communication.
Parents’ messages about retaliatory violence were assessed by both youth and parents using a subset of five items (youth α = 0.77; α = 0.69) from a scale developed by Orpinas et al. (1999). Examples of scenarios presented included, “I would tell my child/your parent tells you”: “If someone hits you, hit them back” or “If you can’t solve the problem by talking, it is best to solve it through fighting.” Youth and parents indicated on a 4-point Likert-type scale whether they agreed or disagreed with statements about what parents tell their children to do when confronted with conflict situations. The questions were reverse coded and summed such that higher scores indicated stronger messages in support of violence.
Analysis
Qualitative
Transcripts were entered into HyperRESEARCH 3.5.2. (ResearchWare, Inc., 2013). A grounded theory approach to data coding was used so that content analysis was inductive (Strauss & Corbin, 1998). Two principal coders with expertise in qualitative research methodology identified themes that emerged from participants’ statements. A coding manual was developed based on the first six interviews and modified as subsequent interviews were coded. Each new theme generated a code and similar codes were grouped thematically.
Joint coding was used to improve trustworthiness and rigor (Strauss & Corbin, 1998). Two coders were responsible for coding each transcript. Meetings among coders and senior researchers were held twice a month to ensure consistency and to resolve coding discrepancies. A systematic review of the text assigned to specific codes was performed halfway through the coding process and any identified adjustments to the coding scheme were implemented, with previously completed transcripts re-coded as necessary.
The dyads’ descriptions of the conflict situations were used to determine whether there was concordance or discrepancy between the parent and early adolescent. Transcripts were analyzed inductively to compare themes within and across interviews for both concordant and discrepant dyads.
Quantitative
Means and frequencies were used to examine demographics. Dyads were coded as discrepant if the youth and parent identified different instigators in one or both conflict situations videos. Multivariable logistic regression models were used to examine the association between attitudes toward retaliatory violence, family communication, and perception of parental attitudes toward violence and discrepancy. Youth gender match with the participating parent, youth age, and parent age were included as covariates in the models to determine the effects of the predictors on discrepancy, controlling for the effects of these correlates. Adjusted odds ratios are reported. Analyses were carried out using Stata 13.1 Statistical Software (StataCorp, 2013).
Results
Demographics
The majority of youth (93%) and all the parents (100%) were African American (Table 1). The average age of the youth was 13.6 (SD = 0.9), and more than half were female (58%). Almost 60% of the parents were the same gender as their child who interviewed and less than half of the parents had received some education post-high school (40%).
Parent and Youth Demographics.
Missing 2 values.
Missing 1 value.
Qualitative Results
Identification of instigators and discrepant dyads
Video 1: Locker scene
There were 75 congruent dyads; 66 dyads identified the “passerby” as the instigator, and 9 dyads identified the “locker dweller” as the instigator (Table 2). The other 26 dyads were discrepant; 21 dyads had a parent identify the “locker dweller” as the instigator while the youth identified the “passerby” as the instigator. The remaining 5 dyads had a parent identify the “passerby” as the instigator while the youth identified the “locker dweller” as the instigator. Although the “passerby” was the first youth to engage physically by bumping into the “locker dweller,” participants’ perceptions of the conflict and who started the fight and why it took place varied. A concordant dyad that included a mother and her son described the conflict in similar terms. The youth said the “passerby” was the instigator because “. . . she shouldn’t have walked past and pushed the other girl’s [‘locker dweller’] books down . . . .” His mother agreed that the “passerby” started the fight and added that she “saw the same thing.” In comparison, a mother in a discrepant dyad who identified the instigator as the “passerby” said,
Well, one of the students [“locker dweller”] [was] putting stuff in the locker. Another one [“passerby”] made a mistake and bumped into the student . . . I don’t think the student [“locker dweller”] should have act the way that she acted. Well, she [“locker dweller”] shouldn’t have said it in a negative way that she did say it. But the other students [“passerby”] also shouldn’t have pushed her because she bumped into her by accident.
Parent and Youth Identification of Instigator: Video 1 Locker Scene and Video 2 Outdoor Tripping Scene (n = 101).
Her child said the “locker dweller” started the fight and described it as, “ . . . instead of asking . . . nicely, ‘Why didn’t you say excuse me?’ she started yelling, and then the other one pushed them.”
Video 2: A group teasing another student
There were 65 congruent dyads; 22 dyads identified the instigator as the “taunter” and 43 dyads identified the “rock thrower” (Table 2). In comparison, there were 36 discrepant dyads; 9 dyads had a parent identify the “taunter” as the instigator while the youth identified the “rock thrower” as the instigator. In the other 27 dyads, the parent identified the “rock thrower” as the instigator while the youth identified the “taunter.” The “rock thrower” was reactive and first to engage physically. When asked to describe the scenario, a youth from a concordant dyad said the “rock thrower” was the instigator, “Because the other girl [‘taunter’] ain’t never hit her. She [‘rock thrower’] swung first . . . Even though she [‘taunter’] was teasing her, she never hit her. Fightin’ is when you’re hitting. Teasing is when you say stuff.” Her mother agreed with the description. In contrast, a mother in a discrepant dyad who identified the “taunter” as the instigator said “. . . the other girl [‘taunter’] started the fight because she came in there picking with the other girls . . . she started it because she was running her mouth.” Her daughter disagreed and believed the “rock thrower” started the fight because she threw the rock.
Quantitative Results
Youth and parents’ mean scores on the attitudes toward retaliatory violence, family communication, and perception of parental messages toward retaliatory violence scales are reported in Table 3. Overall, 47 dyads were concordant, 46 dyads were discrepant on one video, and 8 dyads were discrepant on both videos. Scores on the attitudes toward retaliatory violence scale were compared by instigator choice. With Video 1, parents who identified the “passerby” as the instigator had an average score (M = 12.6, SD = 3) that was significantly higher than parents who identified the “locker dweller” (M = 10.6, SD = 2.7). No significant differences were observed when comparing parents’ choice of instigator in Video 2 and scores on this same scale, or for youths’ choice of instigators in Videos 1 and 2, and scores on this scale. A similar comparison of instigator choice was done for the family communication and perception of parental messages toward retaliatory violence scales, and no significant differences were observed (independent group t-test results not presented in table).
Summary of Parent and Youth Study Variables: Means, Standard Deviations, Ranges, and Cronbach’s α.
Four multivariable logistic regression analyses were conducted to examine the relationships between youth and parent attitudes toward retaliatory violence, family communication, and perception of parental messages about retaliatory violence and discrepancy. The results are presented in Table 4. In the first model, we examined the effects of youth and parent attitudes toward retaliatory violence on discrepancy. The results of the analyses are reported in Table 4, and they revealed a significant effect (aOR = 0.84, 95% CI = [0.72, 0.97]), indicating that as parent scores on the attitudes toward retaliatory violence increased the odds of discrepancy decreased. In the second model, we assessed the effects of youth and parent reports of family communication on discrepancy and no significant effects were obtained. In the third model, we looked at the effects of parental messages about retaliatory violence on discrepancy. A significant effect was found for youth (aOR = 0.84, 95% CI = [0.71, 0.99]), and it indicates that as youths’ perceptions of their parents’ messages supporting retaliatory violence increased, the odds of discrepancy decreased. In the final model, all predictors and covariates were included in the model. We found that the association between the youths’ perceptions of their parents’ messages toward violence was attenuated, while the significant association between parents’ attitudes toward retaliatory violence and discrepancy remained significant (aOR = 0.84, 95% CI = [0.73, 0.99]).
Results of the Logistic Regression Models Examining Associations With Parent and Youth Study Variables and Discrepancy in Identifying Instigators.
Note. Variables included in the models as controls include parent age, youth age, and parent–youth gender match. OR = odds ratio; CI = Confidence interval.
Discussion
This article examined discrepancy in the identification of the instigator of a physical fight and the association between attitudes toward retaliatory violence, family communication, and perceptions of parental attitudes toward violence and discrepancy among a sample of urban youth and their parents. Our results show that parents and youth in our sample did not always view conflict situations the same way. In many cases, parents identified one actor as the instigator while the youth identified the other actor. When discrepancy was examined using quantitative methods, parental attitudes toward retaliatory violence were associated with discrepancy, such that as parents’ attitudes in support of violence increased, the odds of discrepancy decreased. This highlights the importance of parental influence and the parenting challenges associated with communication, advice, and discipline related to conflict in situations where attribution of fault is ambiguous. This may be a particularly important issue in urban environments where exposure to potential conflict situations is likely greater and where parents and youth are likely to hold attitudes in support of retaliatory violence (Orpinas et al., 1999; Solomon et al., 2008). Research examining correlates of discrepancy and potential negative outcomes of these divergent views is needed. In the interim, individuals developing and implementing family-based violence prevention interventions need to be cognizant of the complexity of fault attribution, particularly in urban settings, and design strategies to promote conversations between youth and parents around attribution of blame.
This study focused on discrepancy, yet it should be noted that many dyads agreed on the identification of the instigators. A substantial number of youth, parents, and dyads chose the actors who were first to physically engage in conflict as the instigator. This finding is consistent with research on parents’ responses to bullying, whereby parents are more likely to think that overt forms of aggression are more harmful than relational forms of aggression and, because of this, are more likely to intervene (Mishna, Pepler, & Winer, 2006; Waasdorp, Bradshaw, & Duong, 2011). Parents and youth were more likely to perceive and react to overt acts of violence. In Video 1, two thirds (65%) of the dyads identified the “passerby” or the girl who was the first to engage physically as the instigator, and in Video 2, almost half (43%) of the dyads identified the “rock thrower,” who was also the first to engage physically, as the instigator.
While parent–youth dyad concordance was common, there were a substantial proportion of dyads that were discrepant. In both scenarios, when a dyad was discrepant, the parent was more likely to identify the youth who reacted to the situations as at fault (e.g., the “locker dweller” who reacted to being shouldered; the “rock thrower” who reacted to being laughed at). This finding is interesting in a context where parents and youth are likely to display aggressively biased social information processing (Bradshaw & Garbarino, 2004; Guerra, Huesmann, & Spindler, 2003). It may reflect a finding from our previous qualitative study of these data in which parents expressed a desire for their child to be able to ignore an insult and walk away (Lindstrom Johnson et al., 2013). The fact that the youth were more likely to identify the other aggressor as at fault speaks to the difficulty of this behavior.
In the quantitative analysis, one consistent correlate of discrepancy was parents’ attitudes about retaliatory violence. Parents who held more aggressive attitudes about violence were less likely to be in a discrepant dyad. Although the finding was attenuated in the full model, youth who perceived that their parents’ provided more aggressive messages about retaliatory violence were also less likely to be in a discrepant dyad. Parents’ attitudes about violence are related to the messages that they give their children about conflict resolution and more importantly the messages that their children hear about conflict resolution (Lindstrom Johnson, Finigan, Bradshaw, Haynie, & Cheng, 2011). These findings support the role of parental socialization in influencing youth involvement in violence (Roche et al., 2008). While it is difficult to tell in this study, it may be that aggressive parental attitudes and messages about violence are more congruent with urban adolescents’ attitudes about violence. This congruency may place adolescents at risk of violence involvement. For instance, a study of urban youth found that congruent attitudes supporting violence were associated with an increased likelihood of fighting, weapon carrying, and suspension (Solomon et al., 2008). Our study might suggest that this risk operates through similar perceptions of conflict situations and therefore similar conclusions about the appropriateness of violence.
Social ecological models and SCT suggest that the environmental context, including community exposure to violence, where these behaviors occur are important to consider when deciphering the implications of parental practices and youth experiences. The youth in this study live primarily in urban communities with high rates of interpersonal violence. Research has found that youth living in lower-income urban areas are exposed to more violence than youth living in middle- or upper-income areas (Cooley-Strickland et al., 2009). Studies have reported that more than 80% of children living in urban areas have witnessed community violence (Cooley-Strickland et al., 2009). Witnessing violence may either confirm or contribute to the development of a negatively biased social information processing style or parental perceptions (Bradshaw, Rodgers, Ghandour, & Garbarino, 2009). Youth with negatively biased social information styles are more likely to make hostile attribution biases or infer hostility in others’ motives or objectives (Crick & Dodge, 1994).
Strengths and Limitations
This study is novel in looking at discrepancies between urban parents’ and youth in their perceptions of conflict situations. The novelty is enhanced through the use of videos, in which parents and youth respond to the exact same situation, and the randomization of the order with which fault was designated (i.e., sometimes youth responded first and sometimes parents) strengthens confidence in the findings of the study. In addition, the use of both quantitative and qualitative data allows for a richer story to be told about the complexities of fault attribution. However, the results from this study should be viewed in light of several limitations. First, the association between parental attitudes toward retaliatory violence and dyad discrepancy cannot be generalized beyond this sample. Similarly, situational peer violence was the focus of this study so the results cannot be generalized to the field of teen dating violence or to non-peer violence. Second, the study’s relatively small sample size may have reduced our power to detect a significant influence on dyad discrepancy. Third, the actors in the two conflict videos were all female. Finally, the cross-sectional study design did not allow for us to make causal inferences regarding relationships.
Implications
This study builds upon the existing literature by showing that parents and youth living in urban, high-crime neighborhoods do not always view conflict situations similarly. The findings set the foundation for future research that seeks to examine parent–youth discrepancy related to attribution of blame and conflict. Longitudinal studies are needed to understand attributions and to explore how discrepancy is associated with youths’ involvement in violence as this may be a point of intervention for work focused on reducing youth violence. In addition, research is needed to help us gain a better understanding of the context in which parents and youth are having conversations about violence. The challenge raised in our other study, specifically when urban parents living in high-crime neighborhoods provide inconsistent messages about conflict avoidance, represents a concern that if addressed could lead to reductions in youth involvement in violence (Lindstrom Johnson et al., 2013). Violence prevention researchers must be careful when working with youth and their caregivers to not assume that caregivers always support non-violent approaches to conflict resolution. Communication between parents and their children about non-violent conflict resolution must be specific. Conflict situations are not necessarily black and white, as the videos shown to the study population illustrate. Because of this, messages about conflict resolution and when and if youth should engage in physical fighting may be lost in translation between parents and their children. Parenting or family-based interventions, which are promising approaches to reducing youth involvement in violence, should consider teaching strategies to promote conversations around attribution of blame and non-violent conflict resolution.
Footnotes
Authors’ Note
The content is solely the responsibility of the authors and do not necessarily represent the official views of the funding agency.
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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This publication was supported by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) Contract Number N01-HD-2-3344 and the DC-Baltimore Research Center on Child Health Disparities P20 MD000198 from the National Institute on Minority Health and Health Disparities (TLC).