ollective efficacy and internalizing behavior linkage. The effect of maternal corporal punishment on behavior problems in this sample were not depe ndent on child age, suggesting the harmful influence of maternal corporal punishment is consistent throughout early childhood . Copyright by JULIE MA 2015 This dissertation is dedicated to my family in the U.S. and Korea : Thank you for all your love, support, and encouragement . ................................ ................................ ................................ .......................... ................................ ................................ ................................ ....................... ................................ ................................ ............................. ................................ ................................ ................................ ....... ................................ ................................ ....... ................................ ................................ .. ........................... ................................ ................................ .... ................................ ....... ................................ ................................ ............... ................................ ................................ ................... ...................... ................................ ................................ .................. ................................ .. ................................ ................................ ............................. ................................ ................................ ................................ .. ................................ ................................ ................ ................................ ................................ ................................ ............................. ................................ ................................ ................................ ............. ................................ ............. ............ ................................ ................................ .................... ......................... ................................ ....................... ................................ ................................ ................................ .... ................................ ................................ ................................ ................... ................................ ................................ ........................ ................................ ................................ .................. ................................ ................ . ................................ ................................ ................................ .............................. ................................ ................................ ................................ ....................... ................................ ................................ ................................ ............ ................................ ................................ ................................ ...... ................................ ................................ ................................ ............... ................................ ................................ ................. ................................ ................................ . ................................ ................................ .. ................................ ................................ .............. ................................ .................. ................................ ..................... ................................ ................................ ...................... ................................ ................................ .......... ................................ ................................ ..... ................................ ................................ ..... ................................ ............................... ................................ ....................... ................................ ................................ ................................ ................................ ................................ .... ................................ ................................ ...................... ................................ ................................ ................................ ............................... ................................ ................................ ............................. ................................ ................................ ................................ ................................ ................................ ................................ . ................................ ............................. ................................ .............................. ................................ ................................ ............... ................................ ................................ ................................ ......................... ................ ................................ ................................ ................................ ............ ................................ ............... ................................ ................................ ........ ................................ ................................ ........... ................................ ................................ ................................ ........................ ................................ ................................ ........................ ................................ ................................ ................................ ............................. ................................ ................................ ................................ ............. ....... .................... ........... ................................ ................................ ................................ .............................. ................................ ................................ ............................ ................................ ................................ ................................ ............... ................................ ................................ ................. ................................ ................................ . ................................ ................................ .. ................................ ................................ .............. ................................ .................. ................................ ..................... ................................ ................................ ...................... ................................ ................................ .......... ................................ ................................ ..... ................................ ................................ ..... ................................ ............................... ................................ ....................... ................................ ................................ ................................ ................................ ................................ .... ................................ ................................ ...................... ................................ ................................ ................................ ............................... ................................ ................................ ...... ................................ ........ ................................ ................................ ................................ ........................ ...... ............ ................................ ....... ................................ ........ ................................ ................................ .......... ................................ ................................ ................................ ...................... ................................ ................................ .............. ................................ ................................ ................................ ........................... ................................ ................................ ................................ ........... .......................... .......... ................................ ................................ ................................ ............. ................................ ................................ ......... ................................ .............................. ................................ ................................ ................................ ................................ ............................ ................................ ................................ .......................... ................................ ................................ .............. ................................ ................................ ................................ ............. ................................ ................................ ............... ................................ ............................... ................................ ................................ ................................ ................................ ............ ................................ ................ ................................ ................... ................................ ................................ .................... ................................ ................................ ..... ................................ ................................ ........ ................................ ................................ ... ................................ ................................ ... ................................ ............................. ................................ ..................... ................................ ................................ ................................ ................................ ................................ ................................ ............................. ................................ ............... ................................ ................................ ............................... ................................ ........................... ................................ ............................ ................................ ................................ ................. ................................ ................................ ................................ ....................... ................................ ....................... ................................ ..... ................................ ............................... ................................ .................... . ...................... ................................ ............. ................................ ................................ ................................ ................. ........................ ................................ ................................ ...... ................................ ................................ ......... ................................ ................................ ................................ ...................... ................................ ................................ ............................. ................................ ................................ ................................ ...... ................................ ................................ ............ ................................ ................................ ............... ................................ ................................ ................................ ....................... ................................ ................................ ............ ................................ ................................ ................................ ................................ .............. ................................ ................................ ....... ........................... ............................ ................................ ................................ ................... ........................... ................................ ................................ ..... ......... ........... ................................ ................................ ................... ................................ ................................ .... ............................. .............................. ................................ ................................ ................................ .................... ................................ ................................ ............................. ................................ ........ ................................ ............... ................................ ................................ ......... ..................... ............................... ................................ In detail, existing literature ident ifie s neighborhood dis advantage marked by impoverished economic properties (e.g., neighborhood pove rty and un employment rates), neighborhood - level violence, deteriorated physica l properties of neighborhood environments , and lack of positive neighborhood social processes as significant predictors of a host of , including internalizing and externalizing behavior problems . major public health concern a s they are linked to long - term risk factors throughout the life span including low educational attainment , higher rates of school dropouts , delinquency and drug use , and psychopathology . At the family level, the most significant influence on child development is parental behavior . Parental influence is most salient in early childhood, during which children s development is predominantly dependent on their parents and their exposure to and interactions with social contexts outside the family are directly supervised by parents and/or caregivers. A large tradition of research report s that warm an d supportive parenting , that nurtures trust a nd affection between parent and child, play s a significant role in preventing negative child outcomes . On the contrary , parenting practices that are cont rolling and punitive (i.e. , harsh parenting ) are associated with higher levels of negative child outcomes . Among has received particular attention because of its In addition to the direct influe nces of neighborhood and parenti ng on child behavior problems, a growing number of lite rature suggest s that neighborhoods may also influence children indirect ly by way of their effects on family processes particularly, parenting practices such as the use of corporal punishment . G rounded in the ecological framework that underscores the transactional and reciprocal nature of social contexts , T o better understand the ecologies of child development , t existing scholarship, which The first (Chapter Two) examines the simultaneous direct effects of neighborhood disorganization and m aternal corporal punishment on five - year - internalizing behavior problems. This study also explores the indirect neighborhood influence on child behavior by testing whether maternal corporal punishment is a significant mediator in the neighborhood disorganization and child b ehavior problems associations. Hypotheses are based on the aforementioned conceptual framewor k and prior literature on the roles of neighborhood - and parent - processes on child behavior. The second (Chapter Three) builds on the conceptual and analytic models of the first . Drawing from the conceptual framework of the current disserta tion and perspectives of the risk model and the group differences hypothesis , this chapter explores whether there are discernible differences in the effects of neighborhood disorganization and maternal corporal punishmen t on early behavioral issues by race and ethnicity. Findings of this study are expected to shed light on the intricate relationships between neighborhood - , parent - , and individual child - level processes with race and ethnicity. The third paper (Chapter F our) investigates neighborhood and parent influences on the longitudinal patterns of behavior problems between the ages three to five. Of particular interest in this chapter are: 1) the effects of neighborhood disorganization and maternal corporal punishm ent on between - child differences in initial behavior problems at the mean age of the study sample; 2) the effects of neighborhood disorganization and maternal corporal punishment on the average rate s of change in behavior problems; and 3) whether child age moderate s the effects neighborhood and parenting processes on behavior problems. The final chapter (Chapter Five) of this dissertation provides a summary of the findings from the three empirical . Based on these findings, implications for social work practice and policy are discussed. Although the lit erature has firmly established neighborhood disorganization and parental corporal punishment as determinants of advers e child outcomes, significant gaps remain in this body of scholarship . The foremost of these gaps is the scarcity of a robust, multi - dimensional measure that encompasses the A comprehensive exploration of the relationship between neighborhood and parent ing on child development, as well as the role of parent ing practices in this association , will better explain the mediating mechanism of parenting through which neighborhood influence s early childhood outcomes. knowledge generated by this research can inform the cultural competency of s ocial work practice with parents and children living in disadvantaged neighborhoods . espite the extensive literature on the co nfluence of neighborhood characteristics and parenting practices on negative outcomes during adolescence , there is a paucity of research on how neighborhoods affect pre - adolescen t children . The concentration on in previous neighborhood research could be attributed to the perception that young er children are less affected by ex tra - familial factors compared to adolescents who are more exposed to immediate neighborhood influences . To help address this gap in extant literature, th e current study focuses on the effects of neighborhood disorganization and maternal corporal punishment on the patterns of child behavioral outcomes during the first five years of life . Identifying risk factors in early childhood is particularly crit ical for social work practice, a s problematic behavior that emerges during these earlier years of life may be more amenable to early intervention than problems that are allowed to persist into adolescence or adulthood before being addressed . This chapter aims to address the scarcity of prior research that explored the simultaneous roles of parenting - and neighborhood - process es on problematic behaviors in early childhood with the following objectives : 1) to explore the simultaneous effect s of neighborhood disorganization ( represented by lo w levels of collective efficacy) and maternal corporal punishment on fi ve - year - whether maternal corporal punishment mediates the effect of neighborhood disorganization on child behav ior problems . The sample comprised of 2,472 families in the Fragile Families and Child Wellbeing Study (FFCWS) . T he outcome variables were externalizi ng and internalizing behavior problem s at child age five . T he main predictor variables were neighborhood c ollective efficacy and frequency of maternal corporal punishment. Covariates included i ndividual child - , parent - , and neighborhood - level variables. Multil evel models were employed to investigate the concurrent effects of the predictors on t he outcomes while accounting for the possible clustering of individuals in neighborhoods. T he mediation hypothesis was tested using the product of coefficients approach and bootstrapped confidence intervals for total and specific indirect effects. Matern al corporal punishment, regardless of its frequency, was a significant predictor of higher levels of both externalizing and internalizing behavior problems, even after controlling for the effects of collective efficacy, earlier behavior problem scores, and covariates. The protective influence of collective efficacy on internalizing problems was statistically significant, whereas its effect on externalizing problems was marginally significant, net of other predictors. B ootstrapping replications reveal that the indirect ef fects of collective efficacy on behavior problems through maternal c orporal punishment was not significant, after accounting for the covariates . These findings demonstrate the importance of multilevel prevention and intervention that re duce both neighborhood disorganization and maternal use of corporal punishment for more desirable child outcomes. neighborhood pove rty and un employment rates , deteriorated physica l properties of neighborhood environments , violence and crime rates , and ethnic heterogeneity . More r lack of positive neighborhood social processes to increased as well as including externalizing and internalizing problems . reciprocity and trust between parent and child, o o o o Internalizing Behavior, Age 5 Externalizing Behavior, Age 5 Disrupted Attachment Social Learning Neighborhood Disorganization Theorized Intervening Processes Dependent Variables Maternal Corporal Punishment Note: Child - level covariates: externalizing behavior age 3, internalizing behavior age 3, age, gender Parent - level covariates: age, race/ethnicity , education, family income, relationship status Maternal Warmth Maternal Depression Child - level Covariates Parent - level Covariates Neighborhood Income Collective Efficacy Social Control Social Cohesion and Trust Independent Variables o o Internalizing Behavior, Age 5 Externalizing Behavior, Age 5 Maternal Corporal Punishment Social Control Social Cohesion and Trust Collective Efficacy Maternal Warmth Maternal Depression Child - level Covariates Parent - level Covariates Disrupted Attachment Social Learning Neighborhood Disorganization Independent Variables Theorized Intervening Processes Dependent Variables Family Stress Note: Child - level covariates: externalizing behavior age 3, internalizing behavior age 3, age, gender Parent - level covariates: age, race/ethnicity , education, family income, relationship status Neighborhood Income Fragile Families and Child Wellbeing Study (FFCWS) . Prior to implementing this study, t he Institutional Review Board approved this study as exempt research that as a study that uses secondary data in which respondents cannot be identified, this study is determined exempt from federal regulations . Following IRB approval, application for the Contract Data License that grant ed access to neighborhood and contextual variables in the study (e.g. , census tract information, neighborhood income ) was approved by the Center for Research on Child Wellbeing at Pri n ceton University. The FFCWS is a bi rth cohort study of 4,898 families conducted in 20 large U.S. cities with populations over 200,000. Of these 20 cities, 16 were selected from stratified random sampling based on stratification by policy environments and labor market conditions while the remaining four cities were included because they were of special interest to the fundi ng organization . Mothers were recruited at hospitals in these 20 cities between years 1998 to 2000 at the time of the focal chil s birth. Biological fathers were also recruited either at the hospital whe n their child was born or by telephone. Children born to unmarried parents (about 3,700 children) were purposely oversampled and constitute approximately three - quarters of the full FFCWS sample . This sampl ing design approach was designed to yield an overrepresentation of socioeconomically disadvantaged families. These fragile families and their children were likely to be at greater risk of experiencing adverse socio - economic conditions such as poverty, unstable marital relationships , as well as disadvantaged neighborhood conditions and harsh parenting practices including corporal punishment . The core study consisted of mothers and fathers interviews at focal child s birth (baseline) and were followed up when the child was one year old (Wave 2), three (Wave 3), five (Wave 4), and nine (Wave 5) years of age . Baseline interviews were conducted in - person right after the chil s birth at hospitals and subsequent core studies collected data via telephone surveys. Rich information on at the child - , parent - , family - , and neighborhood - level were ac quired, including demographic characteristics , neighborhood, and parenting behavior that are pertinent to th e current study . Additionally, mothers who participated in the Wave 3 and Wave 4 core interviews were invited to participate in the supplemental In - Home Longitudinal Study of Pre - School Aged Children that collected information on child development and home environment when the chi ldren were three and five years of age. This supplemental in - home s tudy collected data via parent interview and an activity booklet. The parent interview covered multiple domains of information on child s health, nutrition, behavior includin g Achenbach s Child Behavior Checklist (CBCL) the outcome measure s of this study , parenting behavior, discipline, and neighborhood characteristics. Most of the questions in the parent interview were reported by the parent or primary caregiver while the remaining questions were interviewer observations about the home environment, child s appearance , and parent - child interaction. Th e activity booklet in the in - home s tudy was used to collect information on both mothers and children s Home Observation for Measurement of the Environment (HOME) is aternal warmth Data for the current study h ad a hierarchical structure in which families are nested in neighborhoods . Therefore, it is possible that individual s perce ptions of collective efficacy are correlated with perceptions of other residents within the same neighborhood. To effectively account for t h e shared variance in the same neighborhoods, this study employ ed cross - sectional multilevel models . effectively accounting for the hierarchical nature of the data in which families are nested in neighborhoods Model 1 Model 2 Model 3 Dependent V ariables Ext. behavior (or Int. behavior) Ext. behavior (or Int. behavior) Ext. behavior (or Int. behavior) Independent V ariables : Neighborhood Disorganization Collective efficacy Collective efficacy Collective efficacy Independent V ariables : Maternal Corporal Punishment Covariates Ext. behavior , age 3 Maternal corporal p unishment Ext. behavior , age 3 Maternal corporal p unishment Ext. behavior , age 3 Child age Child gender Maternal warmth Maternal depression Race/e thnicity Relationship status Education Mother s age Household i ncome Neighborhood income Model 1 Model 2 Model 3 Dependent V ariables Ext. behavior (or Int. behavior) Ext. behavior (or Int. behavior) Independent V ariables : Neighborhood Disorganization Collective efficacy Collective efficacy Mediator V ariable Covariates Ext. behavior , age 3 Child age Child gender Maternal warmth Maternal depression Ra ce/e thnicity Relationship status Education Mother s age Household i ncome Neigh b orhood income Maternal corporal p unishment Ext. behavior , age 3 Child age Child gender Maternal warmth Maternal depression Race/e thnicity Relationship status Education Mother s age Household i ncome Neighborhood income Maternal corporal p unishment Ext. behavior , age 3 Child age Child gender Maternal warmth Maternal depression Race/e thnicity Relationship status Education Mother s age Household i ncome Neighborhood income 1 1 This study used the ml_mediation command in STATA 13 for multilevel mediation analysis. Variable % Mean (SD) Range n Missing Child outcomes Externalizing behavior, age 5 0.42 (0.25) 0 1.47 2,4 53 0.8 % Internalizing behavior, age 5 0.25 (0.20) 0 1.38 2,4 52 0.8 % Externalizing behavior, age 3 0.65 (0.39) 0 2 2,4 49 0.9% Internalizing behavior, age 3 0.40 (0.24) 0 1.50 2,4 49 0.9 % Neighborhood (dis)organization Collective e fficacy 3.10 (0.65) 1 4 2,4 58 0.6 % Mother's corporal punishment (% ) 0.7 4 (0. 89 ) 0 3 2,4 36 1 .5 % Never 49.8 1,2 14 Only once or twice 3 1.6 7 69 A few times this past month 1 3.1 318 A few times a week or more 5.5 1 35 Mother's depression (%) 0.17 (0.38) 0 1 2,4 70 0 .1 % Yes 17 .2 No 8 2.8 Mother's warmth 0.76 (0.28) 0 1 2,0 18 1 8.4 % Child demographics Age (months) 61.1 1 (2.4 2 ) 57 7 1 2,4 72 0% Sex of focal child (% male) 52 .1 2,4 72 0% Mother's demographics Age (years) 30.2 1 (6.0 1 ) 20 50 2,4 72 0% Race/Ethnicity (%) 2,4 63 0 .4 % White, non - Hispanic 2 1.9 54 0 Black, non - Hispanic 51 .3 1,26 4 Hispanic 2 3.7 5 84 Other 3 .1 7 5 Education (%) 2,4 69 0 .1 % Less than high school 3 2.9 81 3 High school degree or GED 30 .4 75 0 Some college/technical school 25.8 638 College degree or higher 10.9 268 Household income ($) 36,690 (44,057) 0 80,000 2,4 72 Variable % Mean (SD) Range n Missing Relationship status (%) 2,470 0.1% Married 3 0.4 75 0 Cohabiting 1 3.0 320 Not married or cohabiting 5 6.7 1, 400 Neighborhood demographics Neighborhood income ($) 37,806 (18,937) 6,913 157,559 2,471 0% Note : p < .05 or lower are bolded. Model 1 Model 2 Model 3 Fixed Effects Collective efficacy 0.02 3 *** 0.02 1 ** 0.01 3 (0.00 7 ) (0.00 7 ) (0.00 7 ) Externalizing b ehavior, a ge 3 0.345*** 0.326*** 0.30 7 *** (0.011) (0.011) (0.011) Mother's c orporal p unishmen t: never Only once or twice 0.03 7 *** 0.03 5 *** (0.0 10 ) (0.0 10 ) A few times this past month 0.081*** 0.07 6 *** (0.013) (0.01 3 ) A few times a week or more 0.11 6 *** 0.108*** (0.019) (0.019) Maternal warmth 0.0 40 * (0.01 8 ) Mother's depression 0.05 0 *** (0.011) Child demographics Age of child (months) 0.004* (0.00 2 ) Sex of c hild : female 0.01 6 (0.008) Mother's demographics Age of mother (years) 0.00 0 (0.00 1 ) Race/ethnicity : White, non - Hispanic Black, non - Hispanic 0.02 5 * (0.01 2 ) Hispanic 0.01 4 (0.013) Other 0.007 (0.026) Education : l ess than h igh s chool High school degree or GED 0.00 9 (0.01 1 ) Some college/technical school 0.02 5 * (0.01 2 ) College degree or higher 0.037 * (0.019) Relationship Status : married Cohabiting 0.0 18 (0.015) Not married or cohabiting 0.04 0 *** (0.01 1 ) Logged a nnual h ousehold i ncome 0.00 1 (0.003) Logged neighborhood i ncome 0.00 4 (0.010) Model 1 Model 2 Model 3 Constant 0. 270 *** 0. 247 *** 0. 608 *** (0.0 23 ) (0.0 23 ) (0.1 56 ) Random Effects Level - 2 variance ( ) 0.024 ** 0.016 0.018 ( 0.031 ) ( 0.053 ) ( 0.040 ) Level - 1 variance ( ) 0.210 *** 0.208 *** 0.204*** (0.004) ( 0.004 ) (0.0 04 ) Note p < 0. 07 ; * p < 0.05; ** p < 0.01; *** p < 0.001; Standard errors in parentheses Model 1 Model 2 Model 3 Fixed Effects Collective efficacy 0.0 19 *** 0.0 18 ** 0.0 14 * (0.006) (0.006) (0.006) In ternalizing b ehavior, a ge 3 0.3 70 *** 0.3 66 *** 0.3 36 *** (0.01 5 ) (0.01 5 ) (0.01 6 ) Mother's c orporal p unishmen t: never Only once or twice 0.01 2 0.0 16 * (0.008) (0.008) A few times this past month 0.03 0 ** 0.0 35 ** (0.012) (0.01 1 ) A few times a week or more 0.0 32 * 0.0 37 * (0.01 6 ) (0.016) Maternal warmth 0.0 05 (0.01 6 ) Mother's depression 0.060*** (0.010) Child demographics Age of child (months) 0.000 (0.00 1 ) Sex of c hild : female 0.00 4 (0.007) Mother's demographics Age of mother (years) 0.000 (0.001) Race/ethnicity : White, non - Hispanic Black, non - Hispanic 0.0 18 (0.01 1 ) Hispanic 0.0 48 ** * (0.01 1 ) Other 0.02 5 (0.022) Education : l ess than h igh s chool High school degree or GED 0.00 2 (0.0 10 ) Some college/technical school 0.01 6 (0.010) College degree or higher 0.012 (0.016) Relationship Status : married Cohabiting 0.02 6 * (0.013) Not married or cohabiting 0.01 6 (0.0 10 ) Logged a nnual h ousehold i ncome 0.00 3 (0.003) Logged neighborhood i ncome 0.000 (0.0 10 ) Model 1 Model 2 Model 3 Constant 0. 163 *** 0. 152 *** 0.1 46 (0.0 20 ) (0.0 20 ) (0.1 33 ) Random Effects Level - 2 variance ( ) 0.000 0 .000 *** 0.000 (0.000) (0.000) (0.000) 0.179 *** 0.179*** 0.175*** Level - 1 variance ( ) (0.003) (0.003) (0.002) 0. 163 *** 0. 152 *** 0.1 46 Note : * p < 0.05; ** p < 0.01; *** p < 0.001; Standard errors in parentheses Observed Coefficient Bootstrap S.E. z p Percentile 95% C I Lower Upper Model summary for external i zing behavior Indirect effect 0.002 0.001 1.46 0.145 0.00 5 0.00 1 Direct effect 0.015 0.010 1.61 0.108 0.0 41 0.0 0 4 Total effect 0.017 0.010 1.78 0.0 76 0.043 0 .0 05 Model summary for internal i zing behavior Indirect effect 0.001 0.001 1.36 0.174 0.00 3 0.00 0 Direct effect 0.017 0.007 2.41 0.016 0.030 0.003 Total effect 0.018 0.007 2.54 0.011 0.031 0.004 Percentile confidence interval based on 2,000 bootstrap samples. racial/ethnic differences in structural and socio - economic factors. Results demonstrate that the effects of maternal corporal punishment on behavior problems were indistinguishable by race/et hnicity. The significant interaction between collective efficacy and Hispanic race/ethnicity indicate that the protective effects of collective efficacy on internalizing behavior were more pronounced in Hispanic children than white children. Race/ethnici ty, however, did not moderate the associations between collective efficacy and externalizing behavior. Overall, these findings underline the importance of multilevel intervention that strengthens neighborhood collective efficacy and enhances positive pare nting practices to all families r 1. 2. Internalizing Behavior, Age 5 Externalizing Behavior, Age 5 Disrupted Attachment Social Learning Neighborhood Disorganization Theorized Intervening Processes Dependent Variables Maternal Corporal Punishment Note: Child - level covariates: externalizing behavior age 3, internalizing behavior age 3, age, gender Parent - level covariates: age, race/ethnicity , education, family income, relationship status Maternal Warmth Maternal Depression Child - level Covariates Parent - level Covariates Neighborhood Income Collective Efficacy Social Control Social Cohesion and Trust Independent Variables Race and Ethnicity supplemental in - home interviews when the children were 3 and 5 years of age. Total Sample White Black Hispanic Missing Variable ( Range ) N = 2,388 (100%) n = 540 (23%) n = 1,264 (53%) n = 584 (24%) M ( SD ) or % M ( SD ) or % M ( SD ) or % M ( SD ) or % % Child Outcomes Externalizing Behavior, Age 5 (0 1. 5 ) 0.42 (0.25) 0.41 (0.24)* 0.44 (0.26)* 0.42 (0.25)* 0.8% Internalizing Behavior, Age 5 (0 1. 4 ) 0.25 (0.20) 0.22 (0.18)*** 0.24 (0.19)*** 0.30 (0.24)*** 0.8% Externalizing Behavior, Age 3 (0 2 .0 ) 0.65 (0.39) 0.60 (0.35)** 0.67 (0.41)** 0.65 (0.39)** 1.0% Internalizing Behavior, Age 3 ( 0 1.5) 0.39 (0.24) 0.34 (0.20)*** 0.41 (0.24)*** 0.42 (0.25)*** 1.0% Neighborhood Disorganization Collective Efficacy (1 4) 3.10 (0.65) 3.31 (0.56)*** 3.03 (0.68)*** 3.06 (0.62)*** 0.5% Mother's Corporal Punishment (%) 1.5% Never 49% 51%*** 44%*** 59%*** Only once or twice 32% 28%** 35%** 29%** A few times this past month 13% 13%*** 15%*** 9%*** A few times a week or more 5% 8%** 6%** 3%** Mother's Warmth (0 1) 0.76 (0.28) 0.85 (0.24)*** 0.72 (0.30)*** 0.79 (0.26)*** 18.3% Mother's Depression (%) 17% 19%** 18%** 13%** 0.0% Child demographics Age (57 71 months) 61.12 (2.42) 60.34 (2.07)*** 61.32 (2.24)*** 61.40 (2.90)*** 0.0% Sex of focal child (boy) 52% 53% 52% 50% 0.0% Mother's demographics Age (20 50 years) 30.17 (6.00) 32.35 (6.48)*** 29.40 (5.62)*** 29.77 (5.86)*** 0.0% Total Sample White Black Hispanic Missing Variable ( Range ) N = 2,388 (100%) n = 540 (23%) n = 1,264 (53%) n = 584 (24%) M ( SD ) or % M ( SD ) or % M ( SD ) or % M ( SD ) or % % Education (%) 0.1% Less than High School 33% 17%*** 33%*** 48%*** High school degree or GED 31% 23%*** 36%*** 26%*** Some college/technical school 26% 29%*** 26%*** 22%*** College degree or higher 10% 30%*** 5%*** 4%*** Relationship Status (%) 0.1% Married 30% 55%*** 16%*** 36%*** Cohabiting 13% 9%*** 12%*** 18%*** Not married or cohabiting 57% 36%*** 72%*** 46%*** Annual Household Income ($0 800,000) 35,689 (40,695) 61,449 (61,839)*** 27,135 (27,738)*** 30,384 (27,870)*** Neighborhood demographics Neighborhood income ($6,913 157,559) 37,410 (18,579) 50,302 (21,863)*** 30,969 (13,238)*** 39,422 (18,533)*** 0.1% Note: Chi - square tests were conducted for binary and c ategorical variable s. Bonferonni corrected one - way analyses of variance (ANOVA) were conducted for continuous variables. * p < 0.05; ** p < 0.01; *** p < 0.001 Note : p < .05 or lower are bolded. ( = 0.105, p < .01) Model 1 Model 2 Fixed Effects Collective efficacy a 0.014* 0.007 (0.007) (0.017) Race/ethnicity: White, non - Hispanic Black, non - Hispanic 0.028* 0.027 ( 0.013 ) (0.018) Hispanic 0.016 ( 0.014) ( 0.019) Mother's corporal punishment: never Only once or twice 0.034** 0.021 (0.010) (0.021) A few times this past month 0.074*** 0.060* (0.014 ) ( 0.028) A few times a week 0.097*** 0.105** ( 0.020) ( 0.035) Collective efficacy and race interactions Collective Efficacy × Black 0.013 (0.019) Collective Efficacy × Hispanic 0.001 (0.022) Mother's corporal punishment and race interactions Mother's CP (=1) × Black 0.002 (0.025) Mother's CP (=1) × Hispanic 0.046 (0.029) Mother's CP (=2) × Black 0.006 (0.033) Mother's CP (=2) × Hispanic 0.046 (0.042) Mother's CP (=3) × Black 0.027 (0.044) Mother's CP (=3) × Hispanic 0.030 (0.062) External behavior problem, age 3 a 0.308*** 0.307*** (0.012) (0.011) Model 1 Model 2 Maternal warmth a 0.031 0.031 (0.017) (0.017) Mother's depression 0.050*** 0.050*** (0.012) (0.012) Child demographics Age of child (months) a 0.004* 0.004* (0.002) (0.002) Sex of child: girl 0.017* 0.018* (0.009) (0.009) Mother's demographics Age of mother (years) a 0.000 0.000 (0.001) (0.001) Education: less than high school High school degree or GED 0.007 0.007 (0.011) (0.011) Some college/technical school 0.023 (0.012) (0.012) College degree or higher 0.044* 0.046* (0.019) (0.019) Relationship Status: married Cohabiting 0.015 0.016 (0.015) (0.015) Not married or cohabiting 0.040*** 0.040*** (0.011) (0.011) Household income a,b 0.002 0.002 (0.003) (0.003) Neighborhood demographics Neighborhood income a,b 0.008 0.008 (0.010) (0.010) Constant 0.404*** 0.408*** (0.016) (0.018) Random Effects Standard Deviation Level - 2 variance ( ) 0.024* 0.030** (0.036) (0.023) Level - 1 variance ( ) 0.205*** 0.204*** (0.004) (0.004) Note : CP = corporal punishment. a Continuous predictors were grand mean - centered for analyses. b Household income and neighborhood income were log transformed for analyses. p < 0.07; * p < 0.05; ** p < 0.01; *** p < 0.001; Standard errors in parentheses. ( = 0.039, p < .05) in Model 4 p < .05) ( = 0.042, p < .01) Model 3 Model 4 Fixed Effects Collective efficacy a 0.013* 0.002 (0.006) (0.014) Race/ethnicity: White, non - Hispanic Black, non - Hispanic 0.020 0.021 (0.011) (0.015) Hispanic 0.048*** 0.042** (0.012) (0.016) Mother's corporal punishment: never Only once or twice 0.015 (0.008) (0.018) A few times this past month 0.036** 0.028 (0.012) (0.024) A few times a week 0.039* 0.021 (0.017) (0.030) Collective efficacy and race interactions Collective Efficacy × Black 0.003 (0.016) Collective Efficacy × Hispanic 0.039* (0.018) Mother's corporal punishment and race interactions Mother's CP (=1) × Black 0.002 (0.021) Mother's CP (=1) × Hispanic 0.013 (0.025) Mother's CP (=2) × Black 0.013 (0.028) Mother's CP (=2) × Hispanic 0.008 (0.037) Mother's CP (=3) × Black 0.026 (0.037) Mother's CP (=3) × Hispanic 0.034 (0.053) Internalizing behavior problem, age 3 a 0.335*** 0.335*** (0.016) (0.016) Model 3 Model 4 Maternal warmth a 0.006 0.007 (0.016) (0.016) Mother's depression 0.060*** 0.060*** (0.010) (0.010) Child demographics Age of child (months) a 0.000 0.000 (0.002) (0.002) Sex of child: girl 0.007 0.006 (0.007) (0.007) Mother's demographics Age of mother (years) a 0.000 0.000 (0.001) (0.001) Education: less than high school High school degree or GED 0.002 0.003 (0.009) (0.009) At least some college 0.016 0.015 (0.011) (0.011) College degree or higher 0.013 0.011 (0.016) (0.017) Relationship Status: married Cohabiting 0.028* 0.027* (0.013) (0.013) Not married or cohabiting 0.019* 0.019* (0.010) (0.010) Household income a,b 0.003 0.003 (0.003) (0.003) Neighborhood demographics Neighborhood income a,b 0.003 0.004 (0.008) (0.008) Constant 0.217*** 0.218*** (0.013) (0.016) Random Effects Standard Deviation Level - 2 variance ( ) 2.54e - 11*** 3.43e - 12*** (9.88e - 11) (1.04e - 11) Level - 1 variance ( ) 0.175*** 0.175*** (0.003) (0.003) Note : CP = corporal punishment. a Continuous predictors were grand mean - centered for analyses. b Household in come and neighborhood income were log transformed for analyses. p < 0.07 ; * p < 0.05; ** p < 0.01; *** p < 0.001; Standard errors in parentheses. The study results also indicate the universal (i.e., race/ethnicity neutral) effect of mater nal corporal punishment on adverse child outcomes. Professional and clinical settings including pediatric clinics, child care centers, parent education programs, and home visiting programs should continue to inform parents of the growing body of research demonstrating the deleterious influence of maternal corporal punishment on child behavior and provide information on alternative and effective disciplinary strategies. As cultural beliefs and values of minority communities may intricately inform the meani ng - making of corporal punishment to the parents and the children, an awareness and sensitivity to effectively address cultural values associated with parenting practices of minority parents is imperative in these efforts. Another important aspect to cons ider is the intersections of neighborhood, family, and parenting contexts and its complex influence s on child development. To fully consider the ecological and interactive relationships between neighborhood - , family - , and individual - processes, a family - c entered and strength - based approach should be implemented for programs that serve families with children in disadvantaged neighborhoods . , children make more frequent attempts to establish autonomy within their social contexts as they grow older (Brooks - Gunn et al., 1993 ; ; Wadsworth & Santiago, 2008 ). . 1. 2. 3. 4. In - Home Longitudinal Study of Pre - School Aged Children (In - Home study hereafter) 2 2 For example, missing data in Wave 3 maternal warmth was imputed using all study variables from all current (Wave 3) and preceding (Wave 1) interviews, only if the respondent participated in Wave 3 In - Home study during which maternal warmth was assessed. 3 3 The values in the Wave 3 5 - point scales were converted to the values in the Wave 4 4 - point scales as follows: 2 to 1.75, 3 to 2.5, 4 to 3.25, and 5 to 4. Univariate statistics are presented to describe sample characteristics at different time points. T - tests and ANOVA were conducted for time - varying variables to examine whether the differences over time were statistically significant. Correlation analyses were performed to examine the bivariate associations between different study variables, as well as the correlations between time - vary ing variables at baseline and subsequent time points. Lastly, longitudinal multilevel models (MLM, Singer & Willet, 2003), also known as hierarchical linear models (HLM, , were employed to examine the simultaneous effects of neighborhood disorganization and maternal corporal punishment on child externalizing and internalizing behavior problems across ages 3 to 5. Repeated observations over time on the same construct for t he identical person are likely to be correlated. The longitudinal multilevel model can correctly adjust for the within - person variation by estimating repeated measurements of a particular individual as Level 1 units and the individual themselves as Level 2 units. Although participants in this study we re somewhat clustered wi thin the same neighborhoods, prior cross - sectional models of the effects of collective efficacy and maternal corporal punishment on early childhood behavior problems with FFCW S data fo und that the neighborhood - level variation in child behavior problems was not statistically significant after accounting for the same covariates includ ed in this study (see Chapters Two & Three of this dissertation). Therefore, the current study does not i nclude neighborhood as an additional Level 3 unit. The multilevel model is a preferred method for longitudinal data analysis with several advantages. As Singer and Willett (2003) note, the longitudinal MLM does not require balanced data. It effectively uses all available data in model estimation and allows flexibility in regards to missing data across different time points . Another advantage of the multilevel model is that it permits varying time intervals between observations . As such, the multilevel model is an optimal strategy to analyze repeated observations of children in the current study and allows the simultaneous investigation on whether the between - child differences in initial behavior pr oblems as well as the rate of change in behavior problems depend on neighborhood and parenting processes. In this study, repeated observations of children were the Level 1 unit of analysis and the children themselves were the Level 2 unit of analysis. To discern the effects of corporal punishment of infants (age 1) from later corporal punishment (age 3 and age 5), corporal punishment at age 1 was estimated as a time - invariant predictor while ages 3 and 5 corporal punishment was analyzed as a time - varying p redictor. hild age was grand mean centered (mean age = 3.1 years) After the associations between behavior problems and time (child age) were determined using scatter plots, sequentially nested models were developed based on the following steps suggested by Singer and Willett (2003): first, an unconditional model (Model 1 ) examined the average level of behavior problems at mean age and the average annual rate of change in which both the intercepts and time slopes were allowed to vary randomly; next, the conditional model (Model 2) analyzed the effects of between - child vari ability in the main predictors collective efficacy, time - invariant age 1 corporal punishment, and time - varying ages 3 and 5 corporal punishment on mean behavior problems and the linear slope of behavior problems over time. This unconditional model also in cluded a comprehensive set of covariates to examine whether the between - child effects of main predictors on the intercept were affected by the added covariates. As stated earlier, the intercept and time slope were estimated as random effects, which allowe d the intercept (average behavior problem score) and time slope (the effect of age on behavior problem score) to vary across individual children. Neighborhood collective efficacy and maternal corporal punishment were estimated as fixed effects based on th e assumption that the effects of these predictors will not be different across individual children. Finally, the interactions between neighborhood collective efficacy and child age and between corporal punishment and child age were included in the analysi s to test whether the rate of change in behavior problems differed by the main predictors. All analyses were performed with the mixed command in STATA 13 . The analytic sample for the multilevel models included 3,424 respondents (total o f 5,828 observations points) with an average of 1.7 observations per respondent. Specification of the final study model is as follows: Variable Age 1 Age 3 Age 5 p value (M (SD) or %) (M (SD) or %) (M (SD) or %) Child Outcomes Externalizing behavior 0.65 (0.39) 0.54 (0.34) < .001 Internalizing behavior 0.40 (0.24) 0.24 (0.20) < .001 Neighborhood Disorganization Collective efficacy 2.91 (0.70) 3.10 (0.65) < .001 Mother's Corporal Punishment 0.46 (0.86) 0.91 (1.03) 0.71 (0.88) < .001 Never 73% 46% 52% Only once or twice 15% 28% 31% A few times this past month 6% 14% 12% A few times a week or more 6% 12% 5% Child emotionality 3.05 (1.17) N/A Mother's depression (%) 0.16 (0.36) 0.21 (0.41) 0.17 (0.38) < .001 Yes 16% 21% 17% No 84% 79% 83% Mother's warmth 0.88 (0.22) 0.76 (0.28) < .001 Child Demographics Age (months) 15.02 (3.43) 35.61 (2.44) 61.58 (2.70) < .001 Sex of focal child (%) N/A Male 52% Female 48% Mother's Demographics Age (years) 26.29 (6.02) 28.07 (6.04) 30.16 (6.02) 0.954 Race/Ethnicity N/A White, non - Hispanic 21% Black, non - Hispanic 50% Hispanic 26% Other 3% Education N/A Less than high school 34% High school degree or GED 31% Some college/technical school 25% College degree or higher 10% Variable Age 1 Age 3 Age 5 p value (M (SD) or %) (M (SD) or %) (M (SD) or %) Relationship status 2.14 (0.84) 2.18 ( 0.88) 2.25 (0.90) < .001 Married 29% 31% 31% Cohabiting 28% 20% 13% Not married or cohabiting 43% 49% 56% Household income ($) 31,215 (34,920) 34,502 (43,373) 36,662 (42,638) < .001 Neighborhood Demographics Median household income ($) 35,826 (17,036) 36,404 (17,950) 37,847 (18,612) < .001 Note. Chi - square tests were conducted for binary and c ategorical variable s. Two independent samples t - tests or Bonferonni corrected one - way analyses of variance (ANOVA) were conducted for continuous variables. Table 4.2. ( N = 3,705) Note: Time - varying demographic variables (child age, mother's age, relationship status, household income, neighborhood income) are from Wave 3 when the focal child was age 3. p < . 05 or lower are bolded. Table 4.3. Fixed Effects Model 1 Model 2 Externalizing behavior at mean age (intercept) 0.637*** 0.937*** (0.007) (0.125) Rate of c hange in externalizing behavior (slope ) 0.050*** 0.063*** (0.003) (0.017) Collective efficacy 0.037*** (0.008) Corporal punishment, age 1: Never Only once or twice this past month 0.006 (0.014) A few times this past month 0.010 ( 0.0 22 ) A few times a week or more 0.032 ( 0.021) Corporal punishment, age 3 - 5: Never Only once or twice this past month 0.074*** (0.014) A few times this past month 0.106*** ( 0.01 7 ) A few times a week or more 0.147*** ( 0.019) Interactions between main effects and child age Collective efficacy × child age 0.008 (0.005) CP, age s 3 - 5 (Only once or twice) × child age 0.008 (0.008) CP, age s 3 - 5 (A few times past month) × child age 0.004 (0.011) CP, age s 3 - 5 (A few times a week or more) × child age 0.007 (0.013) Emotionality, age 1 0.056*** (0.004) Maternal warmth 0.068*** (0.017) Mother's depression 0.085*** (0.011) Child demographics Sex of child: girl 0.023* (0.010) Table 4.3. Model 1 Model 2 Mother's demographics Age of mother (years) 0.002* (0.001) Race/ethnicity: White, non - Hispanic Black, non - Hispanic 0.039** ( 0.015 ) Hispanic 0.003 ( 0.016) Other 0.016 (0.030) Education: Less than high school High school degree or GED 0.030* (0.013) Some college/Technical School 0.064*** (0.014) College degree or Higher 0.085*** (0.022) Relationship Status: Married Cohabiting 0.024 (0.015) Not married or cohabiting 0.037** (0.012) Logged household income 0.005 (0.004) Neighborhood demographics Logged median household income 0.020 (0.011) Random Effects Standard Deviation Level - 2 variance for intercept ( ) 0.317*** 0.282*** (0.010) (0.012) Level - 2 variance for slope of ) 0.071*** 0.079*** (0.018) (0.019) Level - 1 variance ( ) 0.217*** 0.214*** (0.014) (0.017) Note : N = 5,828 observations (Level 1); N = 3,424 individuals (Level 2). Age of child was grand mean centered. * p < 0.05; ** p < 0.01; *** p < 0.001; Standard errors in parentheses. units per year ( . On average, children living in neighborhoods with higher levels of c ( ( Table 4.4. Fixed Effects Model 1 Model 2 Internalizing behavior at mean age (intercept) 0.388*** 0.595*** (0.004) (0.078) Rate of change in internalizing behavior (slope ) 0.067*** 0.088*** (0.002) (0.011) Collective efficacy 0.026*** (0.005) Corporal punishment, age 1 (baseline): Never Only once or twice this past month 0.011 (0.009) A few times this past month 0.000 ( 0.013) A few times a week or more 0.015 ( 0.013) Corporal punishment, age 3 - 5: Never Only once or twice this past month 0.018* (0.009) A few times this past month 0.013 ( 0.01 1 ) A few times a week or more 0.038** ( 0.012) Interactions between main effects and child age Collective efficacy × child age 0.008* (0.003) CP, age s 3 - 5 (Only once or twice) × child age 0.003 (0.005) CP, age s 3 - 5 (A few times past month) × child age 0.011 (0.007) CP, age s 3 - 5 (A few times a week or more) × child age 0.003 (0.009) Emotionality, age 1 0.023*** (0.003) Maternal warmth 0.030** (0.011) Mother's depression 0.059*** (0.007) Child demographics Sex of child: girl 0.008 (0.006) Table 4.4. Model 1 Model 2 Mother's demographics 0.001 Age of mother (years) (0.001) Race/ethnicity: White, non - Hispanic Black, non - Hispanic 0.008 ( 0.009 ) Hispanic 0.050*** ( 0.010) Other 0.046* (0.018) Education: Less than high school High school degree or GED 0.033*** (0.008) Some college/Technical School 0.077*** (0.009) College degree or Higher 0.076*** (0.014) Relationship Status: Married Cohabiting 0.011 (0.009) Not married or cohabiting (0.008) Logged household income 0.007** (0.002) Neighborhood demographics Logged median household income 0.008 (0.007) Random Effects Standard Deviation Level - 2 variance for intercept ( ) 0.190*** 0.170*** (0.007) (0.008) Level - age ( ) 0.058*** 0.061*** (0.010) (0.010) Level - 1 variance ( ) 0.140*** 0.137*** (0.009) (0.010) Note : N = 5,828 observations (Level 1); N = 3,424 individuals (Level 2). Age of child was grand mean centered. p < 0.07; * p < 0.05; ** p < 0.01; *** p < 0.001; Standard errors in parentheses. This dissertation examined the concurrent influences of neighborhood - and parent - level processes on behavior problems in early childhood, using a diverse sample of children and their mothers living in large U.S. cities. Specifically, the first empirical p aper investigated the direct effects of neighborhood disorganization and maternal corporal punishment on behavior problems as well as the indirect neighborhood effect that is transmitted on child behavior through maternal corporal punishment. The second e mpirical paper explored the extent to which the simultaneous effects of neighborhood disorganization and maternal corporal punishment on child behavior problems vary by race and ethnicity. The third empirical paper examined longitudinal patterns of behavi or problems in early childhood and the joint influences of neighborhood disorganization and maternal corporal punishment during the early years. A comprehensive conceptual framework that draws from the ecological systems perspective, social disorganizatio n theory, social learning theory, attachment theory, the family stress model, and findings from previous empirical literature guided the research questions and hypotheses of each paper . The current chapter will provide a summary of each paper and discuss how the findings of this dissertation contribute to extant theoretical and empirical literature and inform practice and policy. The first empirical paper of this dissertation (Chapter Two) examined the extent to which neighborhood - a nd parent - level processes explain 5 - year - from multilevel models support the proposed hypotheses that both neighborhood disorganization and maternal corporal punishment would predict increased externalizing and internal izing behavior problems. The multilevel approach allowed to effectively control for both within - and between - levels of neighborhood collective efficacy. The direct effe cts of neighborhood and parenting processes on behavior problems held true even after accounting for the b roader parenting contexts and an extensive set of factors at the individual - , family - , and neighborhood - levels. Clearly, the findings of the first pa per demonstrate that early behavior problems are shaped by both proximal and distal social processes. In keeping with these results, the first paper also tested the hypothesis that the effects of neighborhood disorganization on child behavior problems wou ld be partially mediated through maternal corporal punishment. Contrary to expectations, results from multilevel mediation analyses indicate that maternal corporal punishment does not have a statistically significant effect on the link between the neighbo rhood disorganization and behavior problems when individual - , family - , and neighborhood - level factors were accounted for. The second paper (Chapter Three) extended the analytic models of the first paper and examined racial/ethnic differences in the relationship between neighborhood disorganization, maternal use of corporal punishment, and behavior problems among 5 - year - olds. Drawing from extant theoretical and empirical literature, this study anticipated that there would be distinguishable patterns in neighborhood and parent effects between white, Black, and Hispanic children, mainly due to the differences in cultural norms a nd socio - economic conditions by race and ethnicity. Results partially supported the hypothesis that race and ethnicity would significantly moderate the effects of collective efficacy and corporal punishment on child behavior. Explicitly, this study found that the protective influence of collective efficacy on internalizing problems were more pronounced in Hispanic children than white children. Contrary to the study hypothesis, however, the associations between corporal punishment and both externalizing a nd internalizing behavior problems were indistinguishable across the three racial and ethnic groups after controlling for individual - , family - , and neighborhood - level covariates. With regards to racial and ethnic differences in the effects of corporal pun ishment on early behavior problems, the results did not support the hypothesis. Although Black and Hispanic children experienced corporal punishment at disproportionately higher rates than white children, race and ethnicity was not a significant moderator in the corporal punishment and child behavior linkage, net of individual - , family - , and neighborhood - level covariates. The third paper (Chapter Four) explored the simultaneous effects of neighborhood disorganization and maternal corporal punishment on the longitudinal patterns of behavior problems in early childhood. This study also examined whether child age moderated the effects of neigh borhood and parenting processes on early behavior problems. Hypotheses of this study were based on empirical literatu re that underscored the reciprocal associations between behavioral trajectories of children and neighborhood and parenting processes. Findings partially supported the hypotheses. Both neighborhood disorganization and maternal corporal punishment were sig nificant predictors of higher levels of externalizing and internalizing behavior problems at the mean age of the study sample, after considering covariates at the individual - , family - , and neighborhood - levels . Contrary to the study expectations, the effec t of neighborhood disorganization on child internalizing behavior was stronger for younger children, net of other variables in the model, whereas th e association between neighborhood disorganization and child externalizing behavior was not dependent on chi ld age. The effect s of maternal corporal punishment on externalizing and internalizing beh avior problems in this sample were also consistent over time. Grounded in an integrated conceptual framework, the findings of t his dissertation advances current literature concerning the roles of multiple social contexts on early behavior problems. The significant direct effects of neighborhood disorganization and maternal corporal punishment on externali zing and internalizing be havior found support for the tenets of social disorganization theory, social learning theory, and attachment theory. However, maternal corporal punishment was not a significant mediator in the neighborhood disorganization and child behavior association, a finding that runs counter to the family stress model. This suggests that empirically, neighborhood and family processes are separate domains, although a line of literature and theory points to the inter - connected nature between these two social contexts (Berkowitz, 1989; Conger et al., 2000; Simons et al., 1996) . On the whole, the current analyses found support for the link between adverse neighborhood - and parenting - processes and increased behavior problems; yet, the mechanism underlying the relationshi p between neighborhood disorganization and child behavior remains unclear. An important direction for future research is to further explore the individual - and family - level pathways through which neighborhood influences are manifested onto child behavior p roblems using the multilevel framework, in order to facilitate theoretically informed and empirically tested intervention strategies for young children and their parents. Methodologically, the ecological and multilevel approach reflected in the conceptua l framework and analytic models of the current study underlines the scientific value of multilevel research that effectively involves both proximal and distal predictors of child outcomes (Trickett & Beehler, 2013) . The theoretical and empirical bridges a cross multiple levels of social contexts found in this study should be embraced in all aspects of research that aims to identify the ecological associations between social systems and child well - being. Also, the direct effect of neighborhood disorganization on early childhood outcomes, even after controlling for the structural effects of neighborhoods, validates the dynamic influence of collective efficacy as a credible indicator of neighborhood effects. Future literature sho uld focus on including both the structural aspects (e.g., SES, crime, physical disorder) and the social and institutional constructs (e.g., collective efficacy, social network) from multiple sources of data into measures that represent neighborhood charact eristics. All three empirical papers in this dissertation found that early behavioral issues are affected by neighborhood disorganization and maternal corporal punishment even after controlling for confounding influe nces at individual - , family - , and neighborhood - levels. A reflection of the current multilevel analyses in line with the primary mission of the social work profession underscores the value of an ecological and multilevel intervention that promotes social j ustice at multiple contexts. Notably, early childhood experiences of neighborhood disorganization and corporal punishment may become substantial antecedents of social inequality over the life course as children in poverty are at higher risk of experiencin g negative neighborhood conditions and harsh parenting processes (Conger et al., 2000; Ross & Jang, 2000) . In fact, the effect of neighborhood disorganization on internalizing behavior was even stronger for younger children in this study, a compelling evi dence that sheds light on the substantial neighborhood influence at early ages . As such, risk factors across neighborhood and parenting contexts ought to be concurrently articulated in social work practice and policy that aim to support well - being in early years. First, to prevent early behavior problems that are likely to have lasting influences on more serious risk factors throughout childhood, professionals working with parents of young children should facilitate collabor ative partnerships between individual - and family - level programs and key community organizations and stakeholders, with the goal of simultaneously promoting positive neighborhood processes and effective child disciplinary practices. Likewise, investing in direct community services for parents with young children that promote social cohesion and community networks such as parent support groups would be an effective route to prevent early behavioral issues. Evidence - based parent education programs should be made widely available to parents with toddlers and preschool - aged children, during which corporal punishment is most commonly used , to offer parenting support and advocate the use of effective methods of child discipline. These parent ed ucation programs also need to stress the significance of collective monitoring and supervision of young children at the community level to promote positive child outcomes. Moreover, the current mediation analysis found that children living in sociall y disordered neighborhoods are not necessarily at higher risk of experiencing parental corporal punishment. This indicates that practitioners should advocate the use of positive parenting practices to all parents, regardless of their neighborhood conditio ns. Further extrapolation of this finding may also suggest that parent education programs should be made available and accessible in every community, as children in both marginalized and more affluent neighborhoods may be equally vulnerable to experiencin g corporal punishment. In addition, although neighborhood and parenting processes are situated in cultural contexts, the current analyses indicated that overall, the effects of neighborhood disorganization and maternal corporal punishment on early behavioral outcomes are consist ent across white, Black, and Hispanic children. This finding highlights the universal prominence of a supportive neighborhood context and family context on child development. From an ecological and multilevel perspective, this means that intervention eff orts that promote both neighborhood collective efficacy and positive parenting practices would benefit all children regardless of their racial and ethnic backgrounds. However, consistent with prior literature, racial and ethnic differences in structural a nd socio - income, and average household income in neighborhood were apparent in this study. Also, racial and ethnic minority children were overrepresented in disadvantaged neighborhoods and were sub ject to higher rates of corporal punishment. In order to provide relevant resources and culturally - sensitive practices, intervention efforts should be mindful of the racial and ethnic disparities in structural and social contexts. As for the stronger pro tective effects of collective efficacy on internalizing behavior problems of Hispanic children compared to white children, social work practice needs to be designed and tailored to address the unique role of Hispanic/Latino culture that is intricately rela ted to neighborhood processes in Hispanic families and communities . From a policy perspective, the study findings provide scientific support for programs and policy that encourage the development of neighborhood collective efficacy. Resource allocation should give priority to multi - faceted risk assessments that include both structural - and process - oriented neighborhood measure s to ensure that the dynamic processes of neighborhood are reflected on community - based intervention. For instance, providing mor e public resources such as schools, parks, community centers, libraries, and play grounds may directly improve social and environmental conditions and indirectly increase collective efficacy in the most impoverished neighborhoods. The findings i n regards to the effects of maternal corporal punishment also have substantial policy implications. Major human rights treaties and organizations including the stron gly advise to outlaw its use (United Nations Committee on the Rights of the Child, 2006) . Notwithstanding the global initiative that recognizes corporal punishment to be unethical, corporal punishment remains a legitimate, socially accepted, and prevalent childhood experience in the U.S., the only developed nation that has not ratified the Convention on the Rights of the Child to date . The current study suggests that it is time for the U.S. to consider joining the global movement that mandates corporal punishment to be both socially and legally unacceptable, to ensure all children are protected from any forms of violence tha t may put their well - being at risk. BIBLIOGRAPHY