Journal of Clinical Child & Adolescent Psychology, 44(1), 157–168, 2015 Copyright # Taylor & Francis Group, LLC ISSN: 1537-4416 print=1537-4424 online DOI: 10.1080/15374416.2014.895939
WAR AND VIOLENCE
Promoting Adolescent Behavioral Adjustment in Violent Neighborhoods: Supportive Families Can Make a Difference! Lorraine M. McKelvey and Nicola A. Conners-Burrow Department of Family and Preventive Medicine, University of Arkansas for Medical Sciences
Glenn R. Mesman and Joy R. Pemberton Department of Psychiatry, University of Arkansas for Medical Sciences
Patrick H. Casey Department of Pediatrics, University of Arkansas for Medical Sciences
This study examined the moderating effects of family cohesion on the relationship between community violence and child internalizing and externalizing problems at age 18. The study sample consisted of 728 children and families who were part of the Infant Health and Development Program, an intervention study for low-birthweight, preterm infants. Six of eight sites in the Infant Health and Development Program were in large metropolitan areas; two served rural and urban areas. About half of the sample was African American. Research teams collected data from caregivers multiple times in the first 3 years of the target child’s life, and at 4, 5, 6½, 8, and 18 years. Caregivers reported on community violence, neighborhood problems with (a) drug users=sellers; (b) delinquent gangs; and (c) crime, assaults, and burglaries reports when children were 4, 5, and 8 years of age. Family cohesion was assessed twice, at ages 6½ and 8 years, using caregiver reports on the Family Environment Scale. Adolescent self-report of Internalizing and Externalizing Behavior Problems at age 18 were assessed using the Behavior Problems Index. In this study, the association between adolescent psychosocial outcomes and community violence were moderated by family cohesion and gender such that being in a highly cohesive family as a child protected male children from the negative effects of community violence. Findings demonstrate the long-term protective effects of family cohesion on child behavioral development for male children but suggest a need to examine additional supports for females exposed to community violence during childhood.
INTRODUCTION Violence in the home and in the community is a serious threat to children’s safety and well-being and can alter their progression through typical developmental trajectories
Correspondence should be addressed to Lorraine M. McKelvey, Department of Family and Preventive Medicine, University of Arkansas for Medical Sciences, 5301 West Markham #530, Little Rock, AR 72205. E-mail: [email protected]
(Margolin & Gordis, 2000). Given the serious consequences of exposure to violence, it is concerning that findings from a nationally representative sample of children aged 1 to 17 documented that 60.6% of children had witnessed or been victims of violence in their homes, schools, or communities (Finkelhor, Turner, Ormrod, & Hamby, 2009). The link between community violence and the development of psychological distress and behavioral problems in children and adolescents is well established (Fowler, Tompsett, Braciszewski, Jacques-Tiura, & Baltes, 2009;
158
McKELVEY ET AL.
Wilson & Rosenthal, 2003). Multiple studies of children and adolescents have demonstrated a relationship between community violence and internalizing problems, such as depression and anxiety, as well as externalizing problems, such as aggression and conduct disorders. Multiple theoretical models have been used to explain the effect of exposure to community violence on children’s outcomes (Fowler et al., 2009). The Environmental Stress Model (Wandersman & Nation, 1998) posits that community violence effects children’s adjustment through the child’s direct experience of stresses and as stress impacts the family, resulting in negative impacts on interpersonal relationships. It is posited that negative interpersonal interactions, in turn, impact children’s behavioral adjustment through observing and modeling the violence and aggression in their environments (Bandura, 1977) and through effects on children’s developing emotional security (Davies & Cummings, 1994; Maccoby & Jacklin, 1980). Although the literature clearly describes the negative impacts of exposure to violence, we still have much to learn about how experiences have effects in the context of other risk factors, how early experiences impact later development, and how characteristics of the family may exacerbate or ameliorate children’s experiences in the community (Fowler et al., 2009). This study explores these important issues. Family Cohesion Existing research and theory suggests that characteristics of the family have the potential to exacerbate or ameliorate the negative effects of living in a violent community (Plybon & Kliewer, 2001). Although families may help protect children from the impacts of community violence exposure, few studies have directly addressed this question. We chose to explore family cohesion, defined as interactions among family members that are supportive, caring, and affectionate (Moos, 1974), as a protective factor because theories of the impacts of violence on children’s outcomes posit effects through family interactions (Bandura, 1977; Davies & Cummings, 1994), and several studies report an inverse relationship with behavior problems. The association between family cohesion and behavioral problems has been investigated in several different age groups with empirical support for the association demonstrated throughout childhood and into adolescence. Studies of younger children have demonstrated a link between cohesion and behavior problems. In one study, children who had a high level of family cohesion (harmony) during preschool had lower levels of behavioral problems in first grade than did children whose families were less cohesive (Hammes, Aparecida Crepaldi, & Bigras, 2012). In addition, in a study of 823 school-age children, high family cohesion measured at age 6 was
associated with lower internalizing and attention problems at age 11, as rated by both mothers and teachers (Lucia & Breslau, 2006). In a sample of early school-age children, aspects of strong family cohesion (e.g., clear roles and responsibilities, emotional support) were related to lower levels of behavioral problems. Conversely, mothers of school-aged children with behavior problems reported that their family was less cohesive than did mothers of school-aged children without behavior problems (Slee, 1996). In adolescence, researchers have also reported that high family cohesion was related to lower externalizing behaviors (Richmond & Stocker, 2006), and poor family cohesion was found to be related to the diagnoses of oppositional defiant disorder and conduct disorder (Rey, Walter, Plapp, & Denshire, 2000). In addition, low family cohesion was predictive of the frequency and severity of delinquent behaviors (Clark & Shields, 1997; Matherne & Thomas, 2001). Futher, Fosco, Caruthers, and Dishion (2012) found that high family cohesion during adolescence was associated with less emotional distress, less aggression, and better subjective well-being in early adulthood. Associations between family cohesion and internalizing problems have also been reported such that internalizing problems are less likely when high levels of family cohesion are present. For example, family cohesion has found to be predictive of fewer depressive symptoms for adolescent boys (Queen, Stewart, Ehrenreich-May, & Pincus, 2013) and feelings of loneliness for adolescent girls (Johnson, LaVoie, & Mahoney, 2001). These findings have also been confirmed in early adolescents, with low levels of cohesion related to higher internalizing problems (Deng et al., 2006). The literature is clear that the presence of nurturing and caring adults can be a powerful protective factor. It is reasonable to suggest that a close-knit family environment, in which children are able to discuss distressing events they have experienced or witnessed in the community, could help restore a sense of safety for the child and decrease the likelihood of negative outcomes. In a study that closely relates to our research question, Plybon and Kliewer (2001) found, in a sample of 99 children aged 8 to 12, that those in high-risk communities who lived in highly cohesive families demonstrated fewer behavioral problems. We have not identified longitudinal studies beginning in early childhood that have examined how family cohesion interacts with gender and community violence to predict behavioral outcomes into adolescence. Gender Our understanding of how community violence may impact behavioral development differently based on child gender is limited. The aforementioned emotional
FAMILY COHESION AND NEIGHBORHOOD VIOLENCE
security theory does not posit gender differences in the development of behavioral problems (Davies & Cummings, 1994; Maccoby & Jacklin, 1980). However, there are established differences in regulatory capacity between male and female infants (Tronick & Reck, 2009) which suggest the potential for differential development of behavior problems as a result of exposure to violence. There is some evidence from prior research to suggest that the links between community violence and psychosocial outcomes may differ based on child gender, although less is known about this relationship for girls as they have often been excluded from study (Zalot, Jones, Forehand, & Brody, 2007), and the number of studies that examine development across gender are small in number (Fowler et al., 2009). There is evidence to suggest that male children are more likely to be exposed to community violence than female (Farrell & Bruce, 1997; Moffitt, Caspi, Rutter, & Silva, 2001; Singer, Anglin, Song, & Lunghofer, 1995), potentially increasing their vulnerability. However, research is not clear on this point, and findings from a recent meta-analysis of the literature on community violence as it impacts child behavioral outcomes, though limited by the minimal number of studies that included gender, found no moderating effects of gender on the established association (Fowler et al., 2009). However, gender effects have been found in at least two studies examining the moderating effects of family interactions on the relationship between community violence and child outcomes. Forehand and Jones (2003) examined coparent conflict related to parenting as a moderator of community violence on externalizing and internalizing behaviors in a sample of 117 low-income, inner-city African American children ages 8 to 14 (Forehand & Jones, 2003). They found that less than average family conflict protected girls, but not boys, against the effects of community violence with regards to depression and aggressive behaviors. McKelvey and colleagues (2011) examined family conflict in early childhood and gender as moderators of early experiences with community violence on late adolescent outcomes. Results suggested that any experience with conflict or violence, be it in the home or community, placed females at risk for depression and anxiety. For both internalizing and externalizing behaviors, male children in low conflict families were not affected by violence in their communities. For boys, unlike girls, living in a low-conflict home was a protective factor against the effects of the neighborhood. Whereas the overall findings are dissimilar, the predictors in the existing studies were the same, namely, family conflict. In our study, we explore whether a cohesive family decreases the likelihood that children who had early experiences with community violence will exhibit problems with emotions and behavior later in adolescence.
159
Study Goals The lack of a cohesive model of the effect of exposure to community violence on children’s outcomes as well as gaps in the extant literature limit our understanding of the impacts of community violence on child outcomes, especially as it relates to family protective factors and gender differences. There are multiple theoretical models used to describe the effects of the community violence on children’s psychosocial outcomes. Taken together, these theories emphasize that child adjustment in violent communities is altered because of family stress and the greater likelihood of less optimal family interactions, both between parents and children and other family members, as well as through the child’s direct observation, modeling, and expectations of negative interactions. The purpose of the present study is to examine the moderating impact of family cohesion and child gender on the relationship between community risk and adolescent behavioral outcomes, controlling for earlier child adjustment and family context. Specifically, we aim to answer the question, Are the impacts of community violence on psychosocial development universal for all regardless of levels of family cohesion and child’s gender? Using the reviewed models and theories combined with the extant literature, we hypothesize that high levels of family cohesion will be protective for children living with high community violence. We do not make a prediction about whether this will be equally true for boys and girls, as existing theories do not provide guidance (Kroneman, Loeber, & Hipwell, 2004) and the empirical evidence is inconclusive (Forehand & Jones, 2003; Fowler et al., 2009; McKelvey et al., 2011).
METHOD Study Design The Infant Health and Development Program (IHDP; 1990) was a randomized clinical trial evaluating an early childhood intervention (N ¼ 985). The goals of the intervention were to foster the optimal development of children born low-birthweight (LBW; less than or equal to 2,500 g) and preterm (PT; 37 weeks gestational age). Families randomly assigned to the intervention (N ¼ 377) received home visiting services and parent group meetings. In addition to home visits, children in the intervention, from age 1 to 3 years, attended a child development center five days per week. There were eight sites in the IHDP: six located in large metropolitan areas (Boston, MA; Dallas, TX; Miami, FL; South Bronx, NY; Philadelphia, PA; and Seattle, WA) and two served rural and urban areas (Little Rock, AR, and New Haven, CT). The study was approved by the Institutional Review Board of each participating
160
McKELVEY ET AL.
institution. Caregivers provided written informed consent for data collected prior to age 18. At age 18, both caregivers and target children provided data and informed consent. The majority of caregivers were biological mothers, with percentages decreasing as the child aged (range ¼ 97–87%). Retention in the IHDP study for waves from birth to age 8 ranged from 87 to 90% and was 65% for the age 18 wave (McCormick et al., 2006). Attrition analyses indicate that the age 18 participants did not differ from the original sample in most regards, with the exceptions of maternal educational attainment at the time of enrollment, race=ethnicity, and study site (McCormick et al., 2006), which were included as control variables in all analyses. Participants As presented in Table 1, the sample is relatively diverse. About half of the sample was African American, a little more than one third had not completed high school, and half were married. In the original study the sample was stratified with two thirds of the infants born weighing less than 2,000 g and one third weighing between 2,000 and 2,500 g. The existing literature on the outcomes of LBW infants suggest long-term developmental challenges that persist into adolescence. For example, the cognitive performance of children born LBW has been shown to be impacted into adolescence even after controlling for social and family demographics (Whitaker et al., 2006). Given that a child’s appraisals of family and neighborhood and behavior are influenced by TABLE 1 Sample Demographics
cognitive functioning, we excluded children with less optimal cognitive development (those with scores less than two standard deviations below the mean on the Wechsler Intelligence Scale for Children–III (WISC-III; Wechsler, 1991) administered at age 8. The analysis sample included 728 children and their families; 281 received the intervention. Procedures The IHDP research teams collected data from caregivers multiple times in the first 3 years of the target child’s life, and at 4, 5, 6½, 8, and 18 years. Each wave of assessment included caregiver interviews and direct assessments of children’s cognitive and physical development. At the 18-year assessment point, the protocol also included an interview with the target children. Measures The current study employed data from different reporters for covariates (caregiver report and direct child assessment), predictor (caregiver report), and outcome measures (youth report). Table 2 shows correlations and descriptive statistics for predictor and outcome variables. Community violence. Community violence reports were collected when children were 4, 5, and 8 years of age using items developed for the Baltimore study of adolescent parenthood (Furstenberg, Brooks-Gunn, & Morgan, 1987). At each age, the measure was created by summing caregiver reports of neighborhood problems with (a) drug users=sellers; (b) delinquent gangs; and (c) crime, assaults, burglaries. The items were scored on
Total Age of Applicant at Enrollment (M, SD) Child Birth Weight:
WAR AND VIOLENCE
Promoting Adolescent Behavioral Adjustment in Violent Neighborhoods: Supportive Families Can Make a Difference! Lorraine M. McKelvey and Nicola A. Conners-Burrow Department of Family and Preventive Medicine, University of Arkansas for Medical Sciences
Glenn R. Mesman and Joy R. Pemberton Department of Psychiatry, University of Arkansas for Medical Sciences
Patrick H. Casey Department of Pediatrics, University of Arkansas for Medical Sciences
This study examined the moderating effects of family cohesion on the relationship between community violence and child internalizing and externalizing problems at age 18. The study sample consisted of 728 children and families who were part of the Infant Health and Development Program, an intervention study for low-birthweight, preterm infants. Six of eight sites in the Infant Health and Development Program were in large metropolitan areas; two served rural and urban areas. About half of the sample was African American. Research teams collected data from caregivers multiple times in the first 3 years of the target child’s life, and at 4, 5, 6½, 8, and 18 years. Caregivers reported on community violence, neighborhood problems with (a) drug users=sellers; (b) delinquent gangs; and (c) crime, assaults, and burglaries reports when children were 4, 5, and 8 years of age. Family cohesion was assessed twice, at ages 6½ and 8 years, using caregiver reports on the Family Environment Scale. Adolescent self-report of Internalizing and Externalizing Behavior Problems at age 18 were assessed using the Behavior Problems Index. In this study, the association between adolescent psychosocial outcomes and community violence were moderated by family cohesion and gender such that being in a highly cohesive family as a child protected male children from the negative effects of community violence. Findings demonstrate the long-term protective effects of family cohesion on child behavioral development for male children but suggest a need to examine additional supports for females exposed to community violence during childhood.
INTRODUCTION Violence in the home and in the community is a serious threat to children’s safety and well-being and can alter their progression through typical developmental trajectories
Correspondence should be addressed to Lorraine M. McKelvey, Department of Family and Preventive Medicine, University of Arkansas for Medical Sciences, 5301 West Markham #530, Little Rock, AR 72205. E-mail: [email protected]
(Margolin & Gordis, 2000). Given the serious consequences of exposure to violence, it is concerning that findings from a nationally representative sample of children aged 1 to 17 documented that 60.6% of children had witnessed or been victims of violence in their homes, schools, or communities (Finkelhor, Turner, Ormrod, & Hamby, 2009). The link between community violence and the development of psychological distress and behavioral problems in children and adolescents is well established (Fowler, Tompsett, Braciszewski, Jacques-Tiura, & Baltes, 2009;
158
McKELVEY ET AL.
Wilson & Rosenthal, 2003). Multiple studies of children and adolescents have demonstrated a relationship between community violence and internalizing problems, such as depression and anxiety, as well as externalizing problems, such as aggression and conduct disorders. Multiple theoretical models have been used to explain the effect of exposure to community violence on children’s outcomes (Fowler et al., 2009). The Environmental Stress Model (Wandersman & Nation, 1998) posits that community violence effects children’s adjustment through the child’s direct experience of stresses and as stress impacts the family, resulting in negative impacts on interpersonal relationships. It is posited that negative interpersonal interactions, in turn, impact children’s behavioral adjustment through observing and modeling the violence and aggression in their environments (Bandura, 1977) and through effects on children’s developing emotional security (Davies & Cummings, 1994; Maccoby & Jacklin, 1980). Although the literature clearly describes the negative impacts of exposure to violence, we still have much to learn about how experiences have effects in the context of other risk factors, how early experiences impact later development, and how characteristics of the family may exacerbate or ameliorate children’s experiences in the community (Fowler et al., 2009). This study explores these important issues. Family Cohesion Existing research and theory suggests that characteristics of the family have the potential to exacerbate or ameliorate the negative effects of living in a violent community (Plybon & Kliewer, 2001). Although families may help protect children from the impacts of community violence exposure, few studies have directly addressed this question. We chose to explore family cohesion, defined as interactions among family members that are supportive, caring, and affectionate (Moos, 1974), as a protective factor because theories of the impacts of violence on children’s outcomes posit effects through family interactions (Bandura, 1977; Davies & Cummings, 1994), and several studies report an inverse relationship with behavior problems. The association between family cohesion and behavioral problems has been investigated in several different age groups with empirical support for the association demonstrated throughout childhood and into adolescence. Studies of younger children have demonstrated a link between cohesion and behavior problems. In one study, children who had a high level of family cohesion (harmony) during preschool had lower levels of behavioral problems in first grade than did children whose families were less cohesive (Hammes, Aparecida Crepaldi, & Bigras, 2012). In addition, in a study of 823 school-age children, high family cohesion measured at age 6 was
associated with lower internalizing and attention problems at age 11, as rated by both mothers and teachers (Lucia & Breslau, 2006). In a sample of early school-age children, aspects of strong family cohesion (e.g., clear roles and responsibilities, emotional support) were related to lower levels of behavioral problems. Conversely, mothers of school-aged children with behavior problems reported that their family was less cohesive than did mothers of school-aged children without behavior problems (Slee, 1996). In adolescence, researchers have also reported that high family cohesion was related to lower externalizing behaviors (Richmond & Stocker, 2006), and poor family cohesion was found to be related to the diagnoses of oppositional defiant disorder and conduct disorder (Rey, Walter, Plapp, & Denshire, 2000). In addition, low family cohesion was predictive of the frequency and severity of delinquent behaviors (Clark & Shields, 1997; Matherne & Thomas, 2001). Futher, Fosco, Caruthers, and Dishion (2012) found that high family cohesion during adolescence was associated with less emotional distress, less aggression, and better subjective well-being in early adulthood. Associations between family cohesion and internalizing problems have also been reported such that internalizing problems are less likely when high levels of family cohesion are present. For example, family cohesion has found to be predictive of fewer depressive symptoms for adolescent boys (Queen, Stewart, Ehrenreich-May, & Pincus, 2013) and feelings of loneliness for adolescent girls (Johnson, LaVoie, & Mahoney, 2001). These findings have also been confirmed in early adolescents, with low levels of cohesion related to higher internalizing problems (Deng et al., 2006). The literature is clear that the presence of nurturing and caring adults can be a powerful protective factor. It is reasonable to suggest that a close-knit family environment, in which children are able to discuss distressing events they have experienced or witnessed in the community, could help restore a sense of safety for the child and decrease the likelihood of negative outcomes. In a study that closely relates to our research question, Plybon and Kliewer (2001) found, in a sample of 99 children aged 8 to 12, that those in high-risk communities who lived in highly cohesive families demonstrated fewer behavioral problems. We have not identified longitudinal studies beginning in early childhood that have examined how family cohesion interacts with gender and community violence to predict behavioral outcomes into adolescence. Gender Our understanding of how community violence may impact behavioral development differently based on child gender is limited. The aforementioned emotional
FAMILY COHESION AND NEIGHBORHOOD VIOLENCE
security theory does not posit gender differences in the development of behavioral problems (Davies & Cummings, 1994; Maccoby & Jacklin, 1980). However, there are established differences in regulatory capacity between male and female infants (Tronick & Reck, 2009) which suggest the potential for differential development of behavior problems as a result of exposure to violence. There is some evidence from prior research to suggest that the links between community violence and psychosocial outcomes may differ based on child gender, although less is known about this relationship for girls as they have often been excluded from study (Zalot, Jones, Forehand, & Brody, 2007), and the number of studies that examine development across gender are small in number (Fowler et al., 2009). There is evidence to suggest that male children are more likely to be exposed to community violence than female (Farrell & Bruce, 1997; Moffitt, Caspi, Rutter, & Silva, 2001; Singer, Anglin, Song, & Lunghofer, 1995), potentially increasing their vulnerability. However, research is not clear on this point, and findings from a recent meta-analysis of the literature on community violence as it impacts child behavioral outcomes, though limited by the minimal number of studies that included gender, found no moderating effects of gender on the established association (Fowler et al., 2009). However, gender effects have been found in at least two studies examining the moderating effects of family interactions on the relationship between community violence and child outcomes. Forehand and Jones (2003) examined coparent conflict related to parenting as a moderator of community violence on externalizing and internalizing behaviors in a sample of 117 low-income, inner-city African American children ages 8 to 14 (Forehand & Jones, 2003). They found that less than average family conflict protected girls, but not boys, against the effects of community violence with regards to depression and aggressive behaviors. McKelvey and colleagues (2011) examined family conflict in early childhood and gender as moderators of early experiences with community violence on late adolescent outcomes. Results suggested that any experience with conflict or violence, be it in the home or community, placed females at risk for depression and anxiety. For both internalizing and externalizing behaviors, male children in low conflict families were not affected by violence in their communities. For boys, unlike girls, living in a low-conflict home was a protective factor against the effects of the neighborhood. Whereas the overall findings are dissimilar, the predictors in the existing studies were the same, namely, family conflict. In our study, we explore whether a cohesive family decreases the likelihood that children who had early experiences with community violence will exhibit problems with emotions and behavior later in adolescence.
159
Study Goals The lack of a cohesive model of the effect of exposure to community violence on children’s outcomes as well as gaps in the extant literature limit our understanding of the impacts of community violence on child outcomes, especially as it relates to family protective factors and gender differences. There are multiple theoretical models used to describe the effects of the community violence on children’s psychosocial outcomes. Taken together, these theories emphasize that child adjustment in violent communities is altered because of family stress and the greater likelihood of less optimal family interactions, both between parents and children and other family members, as well as through the child’s direct observation, modeling, and expectations of negative interactions. The purpose of the present study is to examine the moderating impact of family cohesion and child gender on the relationship between community risk and adolescent behavioral outcomes, controlling for earlier child adjustment and family context. Specifically, we aim to answer the question, Are the impacts of community violence on psychosocial development universal for all regardless of levels of family cohesion and child’s gender? Using the reviewed models and theories combined with the extant literature, we hypothesize that high levels of family cohesion will be protective for children living with high community violence. We do not make a prediction about whether this will be equally true for boys and girls, as existing theories do not provide guidance (Kroneman, Loeber, & Hipwell, 2004) and the empirical evidence is inconclusive (Forehand & Jones, 2003; Fowler et al., 2009; McKelvey et al., 2011).
METHOD Study Design The Infant Health and Development Program (IHDP; 1990) was a randomized clinical trial evaluating an early childhood intervention (N ¼ 985). The goals of the intervention were to foster the optimal development of children born low-birthweight (LBW; less than or equal to 2,500 g) and preterm (PT; 37 weeks gestational age). Families randomly assigned to the intervention (N ¼ 377) received home visiting services and parent group meetings. In addition to home visits, children in the intervention, from age 1 to 3 years, attended a child development center five days per week. There were eight sites in the IHDP: six located in large metropolitan areas (Boston, MA; Dallas, TX; Miami, FL; South Bronx, NY; Philadelphia, PA; and Seattle, WA) and two served rural and urban areas (Little Rock, AR, and New Haven, CT). The study was approved by the Institutional Review Board of each participating
160
McKELVEY ET AL.
institution. Caregivers provided written informed consent for data collected prior to age 18. At age 18, both caregivers and target children provided data and informed consent. The majority of caregivers were biological mothers, with percentages decreasing as the child aged (range ¼ 97–87%). Retention in the IHDP study for waves from birth to age 8 ranged from 87 to 90% and was 65% for the age 18 wave (McCormick et al., 2006). Attrition analyses indicate that the age 18 participants did not differ from the original sample in most regards, with the exceptions of maternal educational attainment at the time of enrollment, race=ethnicity, and study site (McCormick et al., 2006), which were included as control variables in all analyses. Participants As presented in Table 1, the sample is relatively diverse. About half of the sample was African American, a little more than one third had not completed high school, and half were married. In the original study the sample was stratified with two thirds of the infants born weighing less than 2,000 g and one third weighing between 2,000 and 2,500 g. The existing literature on the outcomes of LBW infants suggest long-term developmental challenges that persist into adolescence. For example, the cognitive performance of children born LBW has been shown to be impacted into adolescence even after controlling for social and family demographics (Whitaker et al., 2006). Given that a child’s appraisals of family and neighborhood and behavior are influenced by TABLE 1 Sample Demographics
cognitive functioning, we excluded children with less optimal cognitive development (those with scores less than two standard deviations below the mean on the Wechsler Intelligence Scale for Children–III (WISC-III; Wechsler, 1991) administered at age 8. The analysis sample included 728 children and their families; 281 received the intervention. Procedures The IHDP research teams collected data from caregivers multiple times in the first 3 years of the target child’s life, and at 4, 5, 6½, 8, and 18 years. Each wave of assessment included caregiver interviews and direct assessments of children’s cognitive and physical development. At the 18-year assessment point, the protocol also included an interview with the target children. Measures The current study employed data from different reporters for covariates (caregiver report and direct child assessment), predictor (caregiver report), and outcome measures (youth report). Table 2 shows correlations and descriptive statistics for predictor and outcome variables. Community violence. Community violence reports were collected when children were 4, 5, and 8 years of age using items developed for the Baltimore study of adolescent parenthood (Furstenberg, Brooks-Gunn, & Morgan, 1987). At each age, the measure was created by summing caregiver reports of neighborhood problems with (a) drug users=sellers; (b) delinquent gangs; and (c) crime, assaults, burglaries. The items were scored on
Total Age of Applicant at Enrollment (M, SD) Child Birth Weight:
