Early Childhood Experiences and Kindergarten Success: A ... › wp-content › uploads › 2015 ›...

School Psychology Review,2005, Volume 34, No. 4, pp. 571-588

Early Childhood Experiences and Kindergarten Success:A Population-Based Study of a Large Urban Setting

John W. Fantuzzo, Heather L. Rouse, Paul A. McDermott,and Yumiko Sekino

University of Pennsylvania

Stephanie Childs and Andrea WeissSchool District of Philadelphia

Abstract. This study examined the unique protective influence of center-based earlycare and education experiences on kindergarten outcomes for children enteringpublic school kindergarten. The 3,969 participants were geographically and de-mographically representative of an entire kindergarten cohort in a large urban schooldistrict. Child age, gender, ethnicity, family poverty, low maternal education, andneighborhood were found to be risks for academic and behavioral adjustment uponkindergarten entry. Controlling for these risks, formal, center-based experienceswere related significantly to higher levels of Language Arts, Mathematics, SocialKnowledge, Motor Skills, and Work Habits performance assessments and atten-dance in kindergarten. Initial advantages associated with center-based early careand education were sustained across the kindergarten year.

Three recent National Research Coun- capacity of early intervention to increase favor-cil reports document the impact of early expe- able outcomes for vulnerable children living inriences on school achievement—Eager to high risk environments. Of particular concernLeam (2001), From Neurons to Neighborhoods are multiple risks faced by children living in pov-(2000), and Preventing Reading Difficulties in erty (Egeland, Carlson, & Sroufe, 1993).Young Children (Snow, Burns, & Griffin, Currently, 40% of children under age 61998). Each report highlights two major reali- live in households with incomes below 200%ties drawn from early childhood research: (a) of poverty (National Center for Children inthe adverse effects of multiple early risk fac- Poverty, 2002). These children are more likelytors on later school functioning, and (b) the than their economically advantaged peers to

Author Note. This research project was supported in part by grants from the William Penn Foundation andthe National Institute of Health and Human Development. Special thanks go to our collaborators at Officeof Early Childhood Education in the School District of Philadelphia. This includes Donna Piekarski, Ad-ministrator; Jennifer Plumer Davis, Director of the Prekindergarten Head Start Program; and ErnestineRedd, Director of Comprehensive Early Learning Center.

Correspondence concerning this article should be addressed to John W. Fantuzzo, The Penn GraduateSchool of Education, University of Pennsylvania, 3700 Walnut Street, Philadelphia, PA 19104-6216: E-mail: [email protected] t

Copyright 2005 by the National Association of School Psychologists, ISSN 0279-6015

571

School Psychology Review, 2005, Volume 34, No, 4

demonstrate early and persistent school diffi-culties, including poor academic performance,grade retention, and emotional/behavioralproblems (Axinn, Duncan, & Thornton, 1997;Duncan, Brooks-Gunn, & Klebanov, 1994),Furthermore, poverty concentrated in largeurban centers is associated with the existenceof multiple health and social environmental riskfactors that have a negative impact on educa-tional and social/emotional outcomes (Coulton& Pandey, 1992; U.S, Department of Educa-tion, 1996). Research has documented signifi-cant relationships between child outcomes andchild and family health factors such as low birthweight, exposure to lead, and maternal mentalhealth (Axinn, Duncan, & Thornton, 1997).Similarly, empirical evidence highlights rela-tions between family social risks, such as pov-erty, maternal education, family literacy, andfamily social support, and children's academicachievement (Christian, Morrison, & Bryant,1998; Duncan, Brooks-Gunn, & Klebanov,1994), Research findings confinn that manyof these factors have a significant negative ef-fect on child development that persists through-out elementary and secondary school (Axinn,Duncan, & Thornton, 1997; Klebanov, Brooks-Gunn, & McCormick, 1994).

A recent population-based, school readi-ness evaluation identified the impact of mul-tiple environmental and social risk factors onthe academic and social development of anentire cohort of children entering public el-ementary school in a large urban center (Weiss& Fantuzzo, 2001). Researchers found thatnearly 80% of children entering first grade hadexperienced at least one major risk factor.Many of these risk factors were evident frombirth, such as exposure to lead, low birthweight, birth to a teen parent, poor prenatalcare, and out of home placements. Findingsindicated that controlling for poverty and age,both environmental and social factors signifi-cantly increased children's risk for academicand behavioral problems as well as high ab-senteeism and grade retention.

Independent studies have found thatphysical and social dimensions of neighbor-hood risk threaten child development. Fami-lies living in poverty often reside in neighbor-

hoods with deleterious housing that createshealth and safety hazards for children (e.g.,exposed lead paint, abandoned properties;Coulton, Korbin, & Su, 1996; Wilson, 1987).A myriad of social problems are also associ-ated with poverty. Density of social problemsin a neighborhood, such as community vio-lence, high rates of teen birth, delinquency, andtruancy, have been found to adversely affectvarious domains of child functioning (Coulton,Korbin, Su, & Chow, 1995; Farver, Natera, &Frocsh, 1999), In accordance with a develop-mental-ecological model, inclusion of thesedimensions will enhance our understanding ofneighborhood risk and child development(Leventhal & Brooks-Gunn, 2003).

In addition to the identification of indi-vidual child and neighborhood risk factors,population-based studies allow researchers toexamine heterogeneous outcomes of vulner-able children to identify strengths. Not all chil-dren exposed to risk factors manifest poor out-comes. Proponents for the study of resilienceexplore the transactional processes betweenindividual characteristics and family or envi-ronmental factors that could account for theseheterogeneous outcomes (Luthar, Cicchetti, &Becker, 2000). This perspective defines resil-ience as a multidimensional construct com-posed of individual and contextual features thathelp explain positive adaptation for childrenliving in adverse circumstances.

Participation in quality early learningexperiences has been identified as one suchfactor that promotes resilient outcomes forvulnerable children (National Research Coun-cil, 2001). Particularly for children from low-income families, participation in center-basedpreschool programs is associated with short-and long-term increases in achievement andschool success (Bamett, 1995). Evaluations ofspecific programs designed to enhance schoolreadiness for young children have demon-strated significant gains in cognitive and so-cial/behavioral outcomes for participants (e.g..Cost, Quality and Outcomes [Peisner-Feinberg,et al,, 2001], the High Scope/Perry PreschoolProject [Schweinhart & Weikart, 1997], andAbcedarian Project [Campbell, Pungelo,Miller-Johnson, Burchinal, & Ramey, 2001]).

572

Early Childhood Experiences and Kindergarten Success

Research further documents that children whoexperience greater risk early in life (such asliving in poverty) evidence greater benefitfrom participation in early learning experiences(Hubbs-Tait et al., 2002),

A prospective study sponsored by theNational Institute of Child Health and HumanDevelopment recently examined the impact ofchild care on early development across 10 na-tionally representative sites (NICHD EarlyChild Care Research Network, 1994). Findingsto date show that overall experience in centercare was associated with better social-emo-tional and cognitive-linguistic outcomes(NICHD Early Child Care Research Network,2000,2002). With respect to the study of chil-dren at high risk for early school difficulties,this study is qualified by its initial exclusioncriteria. Populations of children who were bomto teen mothers, whose family did not speakEnglish, who were placed for adoption, or wholived in a neighborhood "deemed by police toounsafe for visitation" were excluded from studyparticipation (NICHD Early Child Care Re-search Network, 2001a). These groups of chil-dren evidence significant risk factors that haveknown relationships with later school success(Berry, 1992; Brooks-Gunn & Chase-Lansdale,1995), and their exclusion precludes general-ity of the study findings. Population-based re-search is needed to determine if similar pro-tective effects are found for groups of childrenevidencing the full range of risk factors.

The purpose of the present research wasto gain a better understanding of the protec-tive benefits of center-based early care andeducation for a representative sample of kin-dergarten children from a large urban schooldistrict. The relationships between center-based experiences and kindergarten academicachievement and behavioral adjustment wereexamined upon kindergarten entry and acrossthe academic year. Two major research ques-tions were addressed to determine the uniquebenefits of center-based experiences. First,what is the effect of major neighborhood (highdensities of physical and social risk factors)and family context variables on academic andbehavioral difficulties at kindergarten entry?Second, what is the unique contribution of cen-

ter-based early childhood experiences to readi-ness for kindergarten and success in kinder-garten controlling for significant risk factors?This question involved an assessment ofchildren's performance at three points in timeduring the kindergarten year: in the fall, at themidpoint, and at the end of the academic year.We hypothesized that the density of neighbor-hood distress, family poverty, and lower levelof maternal education would constitute a sig-nificant risk for poor skill levels at kindergar-ten entry. Additionally we hypothesized thatcenter-based early childhood experienceswould be associated with enhanced skill lev-els at entry and throughout kindergarten con-trolling for neighborhood and family risks.

Methods

Participants

A representative sample of 3,969 kinder-garten children participated in this study. Chil-dren were equally distributed between males(51 %) and females (49%), with an average ageof 66 months at kindergarten entry. Sixty-fourpercent of the children were African American,17% Caucasian, 15% Hispanic, 4% Asian, and, 1 % Native American', These children com-prised a representative sample of children bomin the urban center enrol led in public school kin-dergarten. They were drawn from kindergartenclassrooms across 172 elementary schools. Allkindergarten classrooms in the school districtoperate for a full day. The participating class-rooms were selected to be representative of thenine geographic school district regions and thedemographics of the entire cohort of 15,343 kin-dergarten students. The demographic breakdownof the entire cohort was 51 % male, 62% AfricanAmerican, 18% Caucasian, 16% Hispanic, 5%Asian, and .1% Native American. The overallparticipation rate was 72%. A linkage betweenthe school district and vital statistics databasefor the city indicated that 78% of the childrenin the cohort were born in the city.

Measures

Early care and education experi'ences. The Early Care and Education Inter-view (ECEI) is a structured parent-teacher in-

573


terview designed to collect information aboutchildren's pre-kindergarten early care and learn-ing experiences (birth to 5 years of age)(Fantuzzo & Cohen, 2003), The interview for-mat consists of a series of questions for childrenentering kindergarten for the first time. Thereare nine specific types of early care and educa-tion experiences derived from the National As-sociation for the Education of Young Children'saccreditation standards and the National Insti-tute of Child Health and Human DevelopmentStudy of Early Child Care (NAEYC, 1998;NICHD Early Child Care Research Network,2001a). These nine types of experiences werecategorized into three main groups: Center-Based care. Informal care, and No Extra care.Center-based care is defined as care provided ina location such as a business or religious pro-gram for more than 10 children, and is often re-ferred to as "center care" (NICHD Early ChildCare Research Network, 2001a), Center-basedcare experiences include nationally funded pro-grams such as Head Start (fully subsidized mul-tigenerational program for low-income childrenaged 3-5 and their parents), locally funded pro-grams such as the school district day care pro-gram (partially subsidized program serving 3-to 5-year-old, low-income children of workingparents who do not meet guidelines for HeadStart) and parent cooperative nursery (partiallysubsidized program serving children from birthto age 5 where parents are required to partici-pate in everyday program operations), and pri-vately run center-based care. Informal care in-cludes three different types of home-based ex-periences: family care, in-home care, and rela-tive care. Family care is defined as care providedfor a small group of children in a private homeby a nonrelative. In-home care is care providedwithin the child's home by a person other than aparent or relative. Relative care is care providedby a related person other than the parent, eitherin the child's or relative's home. The final cat-egory of experiences was called No Extra care,and indicated that the child did not have any ofthe above extra experiences prior to kindergar-ten entry.

Reliability of the ECEI was establishedin three ways. First, each survey was checkedto ensure that it was completed correctly, and

that indications of center-based experiences didnot overlap with indications of No Extra care.Across all 3,969 interviews, 96% were com-pleted correctly. Second, the veracity of paren-tal report information on the parameters of theservices—including the duration of experiences,hours per week of attendance, on-site parent-caregiver contact, and center addresses—wascross-checked with each program's documentedguidelines and parameters. These checks re-vealed an 85% accuracy rate. Finally, a ran-dom sample of interviews were cross-refer-enced with administrative records for knownearly childhood programs, and resulted in aconcordance rate greater than 80%,

Performance assessments. Standard-ized performance assessments were used tomeasure entry level readiness and studentprogress across kindergarten. These assess-ments were routine conducted by all the kin-dergarten teachers in the school district andwere designed to inform curriculum and serveas the basis of parent-teacher conferences.Teachers rated each skill level on a 3-pointLikert scale (1= Improvement Needed, 2 =Making Progress, 3 = Competent). For thisstudy five performance assessments were usedrepresenting distinct dimensions of early child-hood achievement: three cognitive/verbal(Language Arts Skills, Mathematics Skills,Social Knowledge) and two motor/behavioral(Motor Skills and Work Habits). Correlationsamong performance measures range frommoderately low to moderately high (r = ,36 to.73). Language Arts and Mathematic Skillswere correlated at a moderately high level (r =.73) but at a lower level with Motor Skills andWork Habits (r = .48-.51). Motor Skills andWork Habits were correlated at a moderatelevel (r = ,60). Social Knowledge was moder-ately correlated with Language Arts and Math-ematics (r = .51) but correlated at a lower levelwith Motor Skills and Work Habits (r = ,37).Descriptions ofthe measures and their reliabil-ity and validity are presented below.

Language Arts Skills. The LanguageArts Performance Assessment (LAPA) is ateacher evaluation of children's languageachievement. There are 18 items rated on a 3-

574


point scale (1 = Improvement Needed, 2 = Mak-ing Progress, 3 = Competent). These items in-cluded, for example, "understanding print as aform of communication," "recognizes letters,""listens and responds to literature," and "iden-tifies similarities and differences." Perfor-mance scores across all 18 variables were aver-aged to create composite scores. Each compos-ite score was then standardized using the entirecohort of kindergarten language arts performanceassessments (A'= 15,345). Internal consistencywas demonstrated for the LAPA (r = .93, p <.001). Criterion-referenced validity for the LAPAwas established with the Cognitive subscale ofthe Child Observation Record (COR; High/Scope Educational Research Eoundation, 1992;Schweinhart, McNair, Barnes, & Lamer, 1993;r = .36, p < .0001), and the Verbal Subscale ofthe Differential Ability Scales (DAS; The Psy-chological Corporation, 1990; r = .51,/; < .001).The predictive validity of the LAPA was con-firmed by examining its relationship to the sec-ond grade Reading Total score on theTerraNova (TerraNova—Second Edition;McGraw Hill, 1997) (r = .49, p < .0001).

Mathematics skills. The MathematicsPerformance Assessment (MPA) is similar informat to the Language Arts Performance As-sessment, and consists of 17 items rated on thesame 3-point scale. Items include, for example,"matches items one to one," "arranges objectsaccording to size," "recognizes numerals,"and "is aware of the concept of addition asjoining sets." Performance scores across all17 variables were averaged to create com-posite scores and were standardized simi-larly to Language Arts assessments. Inter-nal consistency was demonstrated for theMPA (r = .89, p < .001). Criterion-refer-enced validity for the MPA was establishedwith the Cognitive subscale of the COR (r =.28, p < .001), the Test of Early MathematicsAbility (TEM A; Ginsburg & Baroody, 1990; r= .46, p < .05), and the Verbal subscale of theDAS (r = .33, /7 < .001; note significant thoughlower correlations than with the LAPA). Thepredictive validity of the MPA was confirmedby examining its relationship to the secondgrade Mathematics score on the TerraNova (r= .34,/7<.OOOl).

Motor skills. The Motor Skills Perfor-mance Assessment (MSPA) is a teacher evalu-ation of children's physical development, in-cluding fine and gross motor skills. There are16 items rated on the same 3-point scale. Itemsinclude "runs," "hops," "skips," "holds a writ-ing utensil properly," "cuts on a line," and "tiesshoes." Performance scores across all 16 vari-ables were averaged to create composite scoresand standardized using the entire cohort of kin-dergarten motor performance assessments. In-ternal consistency was demonstrated (r = .89,p < .001), and criterion-referenced validity wasestablished with the Movement and Coordi-nation subscale of the COR (r = .29, p < .001).The MSPA was not significantly correlatedwitb the Verbal subscale of the DAS.

Social knowledge. The Social Knowl-edge Performance Assessment is a teacherevaluation of student's social knowledge. Ex-amples of the nine items, rated on the same 3-point scale, include "knows full name," "knowsfull address," "identifies family members andtheir roles," and "identifies people and places inthe neighborhood." Performance scores acrossall nine variables were averaged to create com-posite scores and standardized using the entirecohort of kindergarten performance assessments.Internal consistency was demonstrated (r = .70,p < .001), and criterion-referenced validity wasestablished with the Social Engagement subscaleof the COR (r =.20, /; < .01). Social Knowledgewas significantly correlated with the VerbalSubscale of the DAS (r=.4\,p< .001).

Work habits. The Work Habits Perfor-mance Assessment (WHPA) is a teacher evalu-ation of children's learning behaviors duringschool-related tasks. There are five items ratedon the same 3-point scale. Sample items in-clude "follows directions," "completes tasksindependently," and "demonstrates attentionspan appropriate to tasks." Performance scoresacross all five indicators were averaged to cre-ate composite scores and standardized usingthe entire cohort of kindergarten performanceassessments (N= 15,345). Internal consistencywas demonstrated for the WHPA (r = .90, p <.001). Criterion-referenced validity for theWHPA was established with the Social Engage-

575


ment subscale of the COR (r = .26, p < ,001),and the Learning Behaviors Scale (LBS; Stott,McDermott, Green, & Francis, 1988) (r= ,48,/?< ,0001), Work Habits was not significantly cor-related with the Verbal subscale of the DAS,

Kindergarten attendance. Children'sattendance was collected for three time peri-ods across the kindergarten year. This variablerepresents the percentage of days that the childwas present out of the total school days in agiven time period.

Family poverty. Family income quali-fication for the Free and Reduced Lunch pro-gram was used as a family indicator of pov-erty. Families qualify for participation in thisprogram if they are receiving TANF publicassistance or provide documentation of low-income status.

Maternal education. Mother's educa-tional level at child's birth was derived fromthe Department of Health's Vital Statisticsdatabase. The database includes the highestlevel of maternal education completed inyears. For the purpose of this study, mother'sfailure to complete high school (< 12 yearsof education) was hypothesized as a risk fac-tor for the child's poor educational outcome.

Neighborhood context. Neighbor-hood constructs were measured using informa-tion from the city's integrated aggregate datasystem for children (KIDS). KIDS contain inte-grated administrative databases aggregatedat the block group level including PublicHealth, Protective Services, and HousingProperty, Census block-group level datawere obtained for the 1,801 block groupscomprising the urban center. Data reductiontechniques with 55 block-level variableswere employed to define empirically "neigh-borhood" while accounting for colinearityamong variables (McDermott, McWayne,Fantuzzo, & Culhane, 2003). A series of ex-ploratory common factor analyses yielded twounique and robust factors representing neigh-borhood context: Social Problems and Struc-tural Danger (r = .81 and ,85, respectively).Standard scores were computed across the1,801 block groups for each dimension. Basedon children's residence block-group, standard

scores for each dimension were used to repre-sent neighborhood context. Neighborhoodswith a high score on the Social Problems di-mension have elevated rates of truancy, childpoverty, teen births, delinquent and dependentout of home placements, and substantiatedabuse and neglect cases. Flevated scores onthe Structural Danger dimension indicateneighborhoods with high density of dangerousproperties, incidences of residence fires, andlead levels that exceed maximal allowances.

School region. Elementary schools inthe large urban school district are distrib-uted equally into nine geographic regions.These regions were used to nest schools (fordata analyses) and account for any regionaldifferences.

Procedures

A representative sample of participantswas drawn from 172 public elementary schoolswithin the large urban school district. Schoolswere selected to be representative of the demo-graphic and economic distributions of each re-gion and the district as a whole. Selected chil-dren within each school were representative ofage, gender, racial, and economic distributionofthe region. This study involved the collectionof three types of data: (a) the Early Care andEducation Interview; (b) kindergarten children'sgender, ethnicity, and block-group level iden-tification; and (c) performance assessments ofchildren's literacy^ numeracy, motor develop-ment, social knowledge, and school attendance.

Prior to the start of the school year, re-search team members and district officials metto draft a confidentiality agreement and datacollection procedures. Subsequently, kin-dergarten teachers were provided with materi-als and instructions for the completion of parentinterviews. Teachers conducted one-on-onestructured interviews with parents of incomingkindergarten children in the first week of school,which is a time set aside for kindergarten teach-ers to spend time in their classrooms before thechildren arrive. Next, the enrollment databasethat includes children's demographic informa-tion and free and reduced lunch status was ob-tained and linked to the information collectedfrom structured interviews by district data

576


managers. Records from each respective data-base were matched according to district-as-signed child identification numbers and veri-fied with child name, teacher and school name,and child birth date. A random sample of linkedrecords was checked for accuracy and foundto be 100%. Finally, teachers collected perfor-mance assessments across three points in timethroughout the kindergarten year. At the endof the school year, this performance databasewas obtained and linked to the existing studysample database by matching child identifica-tion numbers. Block-group density of StructuralDanger and Social Problems contained in theKIDS database were linked to individual childinformation by geocodes within the schooldistrict's database. These geocodes representedthe census block group within which childrenresided when they registered for kindergarten atthe beginning of the school year. Partnership be-tween school district and the city's Departmentof Health permitted researchers to use the birthrecord to identify level of maternal education.Finally, all student identifiers were stripped andthe de-identified, archival dataset was providedto the investigators. These procedures outlinedin the confidentiality agreement were designedto protect fully the privacy of the individualstudent records.

Data Analyses

To determine the relationship betweenrelevant child, family, and neighborhood vari-ables, school regions, early care and educationexperiences, and kindergarten performance andattendance outcomes, multiple logistic regres-sion was conducted. This method was used toassess the status of each independent (explana-tory) variable as a risk or protective factor forlater school success, while simultaneously allow-ing for statistical control of the independent ef-fects of all other predictors. Multiple logisticregression created statistical models based on thecombination of explanatory variables to predictclassification of children into two groups basedon performance outcomes: poor versus ad-equate. Performance outcome variables wereclassified dichotomously based on the sampledistribution. Scores iri the top 85% placed chil-dren in the "adequate" school performance

group, whereas scores in the bottom 15% clas-sified children's outcomes as "poor." This cut-off comports with the second standard deviationof proficiency and provides enough cases on bothsides of the dichotomy to enable sufficient sta-tistical power. Using multiple logistic regression,dichotomous independent and dependent vari-ables allow for the calculation of odds ratios.This statistic is more easily interpretable as thedegree of risk or protection exerted by the inde-pendent variables on each dependent variable(e.g., an odds ratio of 2 means that participatingin a center-based early leaming experience re-sulted in twice the likelihood that a child woulddo well in kindergarten mathematics).

One set of logistic regression analyseswas run using the full sample. A series of sixseparate logistic regression analyses were con-ducted to examine the independent effects ofdemographics and school regions. These analy-ses also examined the independent effects ofearly care and education experiences and theinteraction effects of these experiences andneighborhood context, controlling for the ef-fects of demographic and school region vari-ables. The overall chi-square (score statistics)was examined for each logistic regressionmodel. Significant score statistics indicated thatat least one explanatory variable was statisti-cally significant and thus individual Wald chi-squares were examined. Where significantWald chi-squares were present, odds ratios ofsignificant risk or protective factors were in-spected visually to assess their relative impor-tance for each outcome variable. A sample sizeof 3,969 provides sufficient power to detect asmall effect. With 22 explanatory variables, aminimum sample of 2,136 is needed to detecta small effect (.1) with a = .05 where power =.80 (Faul & Erdfelder, 1992).

To determine the relative significance ofcenter-based early childhood experiences onkindergarten performance across the schoolyear a two-way analysis of covariance wasconducted. The analysis of covariance providesincreased precision by statistically controllingthe sources of variation not directly manipu-lated by the experimenter (Glass, Peckham, &Saunders, 1972). The first factor of theANCOVA had two levels: (a) formal, center-

577

School Psychology Review, 2005, Volume 34, No. 4

based early care and education experiences and(b) other care (informal home-based care andno extra care). The second factor was held asrepeated measures corresponding to three re-porting periods across the school year.Ethnicity, family poverty, and maternal edu-cation were included as covariates to controlfor the significant effects of these variables onchild performance outcomes. A power analy-sis was conducted to determine the necessarysample size to detect a small effect of. 1 wherea = .05 and power is set at .80. This analysisindicated that for two groups we would need asample of at least 788 (Faul & Erdfelder, 1992).

A series of five repeated-measuresANCOVA were conducted representing thefive outcome indicators, and significant maineffects as well as interaction effects were ex-plored using the E statistic. The Bonferronicorrection was used for each individualANCOVA (a = .01 per dependent variable) toreduce the likelihood of making a Type I error.Where significant differences were found,Tukey's post hoc analyses were employed todetermine where individual differences existed.

Results

Distribution of Early Care andEducation Experiences

A series of descriptive analyses wereconducted to determine the distribution of earlycare and education experiences for the sample.Overall, it was found that 21% of children didnot participate in an early care and educationexperience, 64% participated in a center-basedexperience, and 15% participated in an infor-mal home-based experience. Gender and eth-nic differences in experiences were tested us-ing the chi-square statistic. There were no sig-nificant gender differences in children's expe-riences (x' (2, A = 3,969) = 2.23, p = .327).There were, however, significant ethnic differ-ences (x^{m,N^ 3,969) = 260.09, p < .0001).White and African American children experi-enced greater proportions of center-based care(67% and 69%, respectively), and Hispanic andAsian children were more likely to have noextra experiences (41% and 56%, respectively)prior to kindergarten entry.

School Region, Child Demographic,Family Context, and NeighborhoodContext Variables

Multiple logistic regression tests of theeffects of schools nested within school regionsdid not indicate a significant impact on schooloutcomes. Therefore, school region was not in-cluded in the final regression models. Table 1shows the bi-univariate relationships betweenchild demographics and family and neighbor-hood context variables. Although both the Struc-tural Danger and Social Problems dimensionsof neighborhood risk were significantly corre-lated with the outcomes, when considered simul-taneously in a multiple logistic regression modelwith demographic and family context variablesthey did not contribute unique variance. TheScore statistic indicated that demographic andfamily context variables were significantly re-lated to children's kindergarten readiness out-comes across all regression models. Table 2 pre-sents the odds ratios and probability levels forthe independent effects of each demographic andfamily context factor on each performance as-sessment score at the beginning of kindergar-ten. Children's age, gender, ethnicity, familypoverty, and maternal education were signifi-cantly related to Language Arts performance atthe beginning of kindergarten. Score (9, A'= 3,328) = 130.88, p < .0001). Younger chil-dren, boys, students of African American orHispanic ethnic backgrounds, children whoqualify for free/reduced lunch, and childrenwhose mother did not complete 12th gradewere at greater risk for poor Language Artsoutcomes in the beginning of kindergarten. ForMathematics, children's age, gender, ethnicity,family poverty, and maternal education weresignificant predictors. Score (9, A' - 3325) -116.53, p < .0001. Similar to Language Arts,younger children, boys, African American andHispanic children, children who qualify forfree/reduced lunch, and children whose motherdid not complete 12th grade were more at riskfor negative outcomes. In relation to children'smotor skills, age, gender, family poverty, andmaternal education were significant risk fac-tors and being African American or Asian wasa significant protective factor. Score (9, A' =3,629)= 135.25,/?<.0001.

578


Table 1Correlations Between Child, Family, and Neighborhood Variables, and

Performance Outcomes

Variables

Child Age

Family Poverty

Maternal Education

Neighborhood Structural Danger

Neighborhood Social Stress

Note. **p<.0001.*p<.01.

LanguageArts

.06**

.15**

.21**

-.16**

-.18**

Performance Assessments

Mathematics

.07**

.13**

.19**

-.15**

-.17**

SocialAwareness

.02

08**

,.16**

-.09**

-.12**

•

MotorSkills

.06**

.07**

.13**

-.06**

-.07**

WorkHabits

.04*

08**

.11**

-.08**

-.07**

Similar patterns of results were found forsocial and behavioral readiness outcomes. Gen-der, ethnicity, family poverty, and maternal edu-cation were significant dsk factors for poor so-cial knowledge outcomes, where boys and chil-dren from Hispanic or Asian backgrounds, chil-dren who qualify for free/reduced lunch, andchildren whose mother did not complete 12thgrade were at a significantly greater risk thanother children. Score (9, A'= 3,520) = 91.33,/? <.0001. Significant risk factors for poor work hab-its were child age, gender, ethnicity, and mater-nal education. Score (9, A'= 3,645) = 99.44,/? <.0001. For this performance assessment, youngerchildren, boys, Hispanic children, and childrenwhose mother did not complete 12th grade weresignificantly more at risk for poor work habits.Both risk and protective factors were found forchildren's attendance at the beginning of kin-dergarten. Score (9, A' = 3,688) = 103.33, p <.0001. Asian children were more than twice aslikely to demonstrate positive attendance,whereas children living in families who qualifyfor free/reduced lunch or whose maternal edu-cation level is below 12th grade were signifi-cantly more likely to have poor attendance.

Early Care and Education Experiencesand Early School Success

The score statistic was significant for allmultiple logistic regression models evaluating

the effects of early care and education experi-ences accounting for significant demographicand neighborhood variables. Table 2 presentsthe odds ratios and probability levels for theindependent effects of early care and educa-tion on early school success. Early care andeducation experiences were significantly re-lated to children's performance assessmentsin Language Arts. It was found that childrenwho participated in center-based experi-ences were significantly more likely thanchildren with no extra care to do well inLanguageArts, S (11, A = 3,328)= 170.71,p < .0001. Furthermore, children who hadinformal home-based experiences were signifi-cantly more likely than children with no extracare to do well in Language Arts. Similar re-sults were found for mathematics, where chil-dren with center-based or Informal experienceswere significantly more likely to do well thanchildren with no extra care, S (11, A' = 3,325)= 160.58,/?<.0001.

For each of the social and behavioraloutcomes no significant findings were discov-ered for children who had informal experi-ences. Children with center-based experi-ences were significantly more likely to dem-onstrate advanced motor skills, S (11, A' =3,629) = 140.41, /7 < .0001, social knowledge,S (11, A = 3,520) = 105.34, p < .0001, andhave better attendance, S (11, A' = 3,688) =124.03, p < .0001. Specific tests of the inter-

579


Table 2Odds Ratios and Probability Levels for Demographic, Neighborhood, and

Early Care and Education Experience variables on Positive ScboolOutcomes at tbe Beginning of Kindergarten

Risk/ProtectiveFactor

Gender (boys)

Age

African American

Hispanic

Asian/Other

Family Poverty

Maternal Education

Structural Danger

Social Stress

Center-based Care

Informal Care

LanguageArts

(N=3328)

.65****

1.06****

.66*

50***

.68

.75**

61****

.79

.92

2.02****

1.62**

Mathematics(Af=3325)

.68***

1.06****

.68*

50***

.94

.75**

.60****

.89

.92

2.16****

1.78**

Outcome Variable

MotorSkills

(A? =3629)

50****

1.08****

1.85****

1.10

299**

.81*

.68***

.88

.84

1.61****

1.38

SocialAwareness(Af=3520)

.78*

1.03

1.02

.55***

.55*

.74**

.70***

.89

.99

1.43**

1.13

WorkHabits

(A? =3645)

50****

1.05****

.75*

.71

1.45

.88

.64****

.88

1.14

1.18

1.30

Attendance(A? =3688)

.95

1.02

1.15

1.07

2.69*

52****

.58****

1.09

.80

1.56****

1.30

*p<.05. **p<.0\. ***p<.001. ****p<.0001.Note. Significance is based on Wald chi-square statistics. An odds ratio less than 1 indicates a risk factor for positiveschool outcomes, whereas an odds ratio greater than 1 indicates a protective factor.

actions between center-based early care andeducation experiences and neighborhood con-text revealed no significant interactions.

Effects of Early Care and EducationExperiences Across Time

Two-way ANCOVAs were conducted toanalyze the effects of center-based early care andeducation across time, controlling for significantchild and family variables. Standardized perfor-mance assessments were applied as the repeatedmeasure to evaluate Language Arts, Mathemat-ics, Motor Skills, Social Knowledge, and WorkHabits outcomes across the kindergarten year.Table 3 presents mean scores and significant

group differences for each standardized perfor-mance assessment across time.

For Language Arts, significant main ef-fects were found for early care and educationexperiences, F(l, 3146) = 71.45,p<.0001 andtime, F{2, 3145) = 2264.73, p < .0001. Acrossall three points in time, children with center-based experiences scored significantly greaterthan children without these experiences (effectsize = .31, .29, and .25, respectively). Further-more, Language Arts scores at the end of theyear (Time 3) were significantly greater thanscores in the middle of the year (Time 2), andmidyear scores were significantly greater thanbeginning of the year scores (Time 1). An

580


Table 3Mean Kindergarten Performance Assessments: Center'Based

Care and Education Across Tune

Standard Scores

Performance Area Time Time 2 Time 3

Language Arts (A' = 3,151)

Center-based care 49.88(9.52) 55.03(9.13) 58.99(8.72)

Othercare 46.88(10.13) 52.26(10.06) 56.63(10.23)

Mathematics (N = 3,180)


Othercare 47.24(10.55) 51.04(9.05) 54.47(8.80)

Motor Skills (N= 3,418)


Othercare 47.43(9.96) 54.10(8.88) 57.69(7.58)

Social Knowledge (A'= 3,294)


Othercare 47.47(9.92) 52.15(8.50) 54.76(7.72)

Work Habits (N= 3,465)


Othercare 48.64(9.34) 51.33(9.02) 52.80(8.51)

Note. Numbers in parentheses represent standard deviations.

identical pattern of results was found for Math-ematics. Significant main effects were foundfor early care and education experience, F{ 1,3175) = 45.51, p < .0001 (effect size = .25,.20, and .20, respectively), and time, F(2,3174)= 1276.57,/?<.0001.

Likewise, the ANCOVAs for SocialKnowledge and Motor Skills showed the su-periority of center-based care and the positiveeffect of kindergarten experience. Significantmain effects for children's motor skills werefound for type of early care and education ex-perience, F(], 3413) = 34.5 l ,p< .0001. Across

all three points in time, children with center-based experiences scored significantly greateron Motor Skills than children without center-based experiences (effect size =. 19,. 19, and. 18,respectively). Significant main effects were alsofound for time, F(2,3,412) = 2448.07, p < .0001.For Social Knowledge, significant main effectswere found for type of early care and education,E{\, 3,289) = 65.09, p < .0001 and time, F{1,3,288) = 1236.83, p < .0001. Across all threepoints in time, children with center-based ex-periences scored significantly greater on so-cial knowledge than children without center-

581


based experiences (effect size = .25, .23, and.27, respectively). Additionally, performancewas significantly greater at Time 3 than at Time2, and Time 2 motor skills were significantlygreater than skills at Time 1.

Children's work habits significantly in-creased across kindergarten, f(2, 3,459) =607.27, p < .0001; however, there was nosignificant main effect for care and educationexperience. Work habits scores at the end ofthe year (Time 3) were significantly greaterthan scores in the middle of the year (Time 2),and mid-year scores were significantly greaterthan beginning of the year scores (Time 1).

Discussion

The primary objective of this researchwas to conduct a population-based investiga-tion of the unique protective influence of cen-ter-based early care and education experienceson kindergarten outcomes. To fulfill this ob-jective, it was important to understand howschool region, neighborhood context, fam-ily context, and child demographic variablesrelated to academic and behavioral out-comes at kindergarten entry. The unique in-fluence of children's center-based experi-ences upon kindergarten entry and acrosskindergarten was then explored by control-ling for these variables.

Demographic and Context Variables

The first assessment, early in the kinder-garten year, revealed overall that younger kin-dergarten children and boys were more likelyto evidence lower levels of academic and be-havioral competencies than older children andgirls. This supports previous literature docu-menting systematic age and gender differencesin young children's cognitive and behavioralschool outcomes (Holahan et al., 2002; Kurdek& Sinclair, 2001; NICHD Early Child CareResearch Network, 2002). With respect to ra-cial/ethnic minority differences, AfricanAmerican, Hispanic, and Asian children weremore likely to show lower levels of languagearts early in the kindergarten year than Cauca-sian children. African American and Hispanicchildren also demonstrated early risks in math-ematics and work habits skills as compeired to

Caucasian children. These findings comportwith recent national studies of kindergartenperformance showing lower school outcomesfor African American and Hispanic childrencompared to Caucasian (West, Denton, &Germino Hausken, 2000). In addition to dem-onstrating lower levels of language arts skills,Asian and Hispanic children evidenced lowerlevels of social knowledge. Because this di-mension captures children's ability to verbal-ize personal knowledge of their social envi-ronment (e.g., knowledge of home address, roleof family members), it is likely that childrenwhose native language is not English wouldhave increased difficulty. In the present study,African American children showed a signifi-cantly higher level of motor skill developmentthan the Caucasian children early in the kin-dergarten year. West and his colleagues (2000)also found similar results. Asian children dem-onstrated high levels of attendance. This find-ing is consistent with literature that highlightsthe cultural value of school participation forAsian families (Steinberg, 1996).

Findings from this study support the hy-pothesis that specific factors within the familycontext significantly account for unique vari-ance in children's early school success. Bothfamily poverty and low maternal education(less than high school diploma) were signifi-cant risk factors for poor academic and behav-ioral outcomes controlling for child demo-graphics and neighborhood risk factors. Thesedata support the general research underscor-ing the proximal infiuence of the family con-text and the specific research on the impact ofpoverty and maternal education and earlyschool success (Axinn et al., 1997; Christianet al., 1998; Duncan et al., 1994).

Empirically derived dimensions ofneighborhood risk were significantly correlatedwith outcomes. However, they did not accountfor a unique amount of variance in academicand social outcomes in our multivariate mod-els. Family poverty and maternal educationwere dominant in the model and showedunique relationship with outcomes. Significantrelationships between these family and neigh-borhood risk factors (significant correlationsranged from .20-.41) indicated that family

582


poverty and low levels of maternal educationwere highly associated with living in neigh-borhoods with the highest density of structuraldanger and social problenns.

Although the neighborhood dimensionsdid not account for unique variance in ourmultivariate model, the current study extendsthe neighborhood risk literature by applyinglatent structures of neighborhood risk factors,derived from an entire urban population, tounderstand the influence of neighborhood riskon kindergarten performance. This inquiry en-abled researchers to examine relationships be-tween a broad range of contextual risk factorsevident across an entire population. Further-more, whereas previous literature has exam-ined specific individual neighborhood risk vari-ables that have known relationships with oneanother (e.g., poverty and crime), the latentdimensions utilized in this study accountedfor colinearity among an array of 55 block-level neighborhood risk variables trackedroutinely by municipal governments. Thesetechniques allowed researchers to accountfor distinct dimensions of neighborhood riskassociated with living in a large urban centerand to examine how these dimensions relateto individual child outcomes in a complexmultivariate model include child and familycontext variables.

Early Care and Education Experiences

For the children entering kindergarten ina large, urban public school district, nearly two-thirds had formal, center-based child care ex-perience prior to kindergarten (64%). Approxi-mately one of five had no extra child care ex-periences (21 %) and 15% had informal, home-based experiences. These prevalence figuresare supported by the Early Childhood Longi-tudinal Study of kindergarten children con-ducted by National Center for Educational Sta-tistics. The kindergarten data, involving a na-tionally representative sample of kindergar-ten children, showed that 69% of the childrenentering kindergarten had a formal, center-based experience prior to kindergarten (Westet al., 2000). In the present study, Hispanic andAsian families evidenced a different pattern ofchild care use than Caucasian and African

American families. The Hispanic and Asianfamilies tended to be less likely to utilize cen-ter-based child care and more likely to keeptheir children at home. In the urban settingwhere this research was conducted, the find-ings may reflect the limited language capacityof center-based programs for Hispanic andAsian families and issues related to recent im-migration to the urban setting.

Controlling for demographic, family, andneighborhood risk factors, this study demon-strated that children who participated in for-mal, center-based early childhood programsprior to public school entry performed at higherlevels than their peers across all academic andsocial areas in kindergarten. At kindergartenentry, children with center-based experiencesperformed significantly better than their peerswith other care experiences (informal experi-ences or no extra child care experiences) inlanguage arts, mathematics, motor skills, so-cial knowledge, and attendance. In addition toadvantages at the beginning of the year, chil-dren who participated in center-based earlyleaming experiences retained their advantagethroughout the kindergarten year. At the endof kindergarten, children with center-basedearly learning experiences evidenced signifi-cantly greater literacy, mathematics, motorskills, and social knowledge outcomes thantheir peers who had other care experiences.

These findings support existing literaturethat has documented the impact of center-basedearly childhood programs on children's aca-demic outcomes. Model early intervention pro-grams, such as High/Scope Perry PreschoolProject and the Carolina Abcedarian Project,have demonstrated significant improvement onchildren's early reading and mathematicsachievement while in preschool (Campbell &Ramey, 1994; Schweinhart & Weikart, 1997).Others have documented increases in generalknowledge, verbal intelligence, vocabulary,and mathematics (Burchinal, Ramey, Reid, &Jaccard, 1995; Christian et al., 1998).

This research also supports previousfindings that have established relationshipsbetween children's early experiences and latersocial and behavioral functioning. Studies havedocumented an increase in positive class-

583


room behavior, social engagement, and co-operation with peers for children who par-ticipate in center-based early childhood pro-grams (Burchinal et al., 1995; Christian etal., 1998; Clarke-Stewart, 1991; Creps &Vernon-Feagans, 1999). Similarly, childrenwho participate in center-based experienceshave demonstrated greater complexity and so-cial competence during play with peers (Aureli& Colecchia, 1996).

Similar to preschool findings, longitu-dinal research has emphasized the long-termbenefit of center-based child care experienceson children's later academic and social suc-cess (Broberg, Wessels, Lamb, & Hwang,1997; Burchinal, Campbell, Bryant, Wasik, &Ramey, 1997), Bamett (1995) provided a re-view of longitudinal research examining theeffects of model demonstration projects andlarge-scale public programs designed for chil-dren from economically disadvantaged fami-lies. He found that participation in such pro-grams can produce sizable effects on laterschool achievement, grade retention, place-ment in special education, and social adjust-ment up to 15 years after program participa-tion. Longitudinal follow-up of children whoparticipated in the Abcedarian program dem-onstrated increased reading and mathematicsachievement that extended through age 21(Campbell et al., 2001). Other research hasdocumented that early experience in child carepredicted gains in social competence immedi-ately following program participation that re-mained relatively stable from preschoolthrough age 15 (Campbell, Lamb, & Hwang,2000).

Whereas immediate and long-term ef-fects have been established for a variety ofspecific intervention programs such as thosecited above (e,g. Campbell & Ramey, 1994;Schweinhart & Weikart, 1997), much less isknown about the positive benefits across a va-riety of early experiences and for populationsof children experiencing a full range of riskfactors. In fact, Barnett (1995) reports thatamong the studies in his review, only theAbcedarian project explored the effects ofparticipation in a variety of child care ex-periences. Researchers found that compari-

son group children who participated in com-munity-based programs that met federalguidelines for quality had higher IQ scores thanchildren who did not, though the effect sizewas half that of the Abcedarian program(Burchinal, Lee, & Ramey, 1989).

Since Bamett's review (1995), some re-cent national studies have highlighted the valueof formal, center-based experiences and there-fore, lend support to the findings ofthe presentstudy. Findings from the Early Childhood Lon-gitudinal study have indicated that childrenwho had experienced center-based care priorto kindergarten were more likely to scorehigher on reading and mathematics assess-ments than children who did not have theseexperiences (Innes, Denton, & West, 2001).The NICHD Study of Early Child Care hasevaluated the effects of a range of child-careexperiences. This study examined parent'snatural selection of child care experiences andwas therefore able to examine individual out-comes for children across a variety of earlyexperiences. Thus far, findings from this studyhave reported child outcomes up to 4.5 yearsof age. Overall, researchers have found thatchildren who participate in center-based earlylearning experiences demonstrate significantlybetter cognitive and language skills by kinder-garten entry as compared to children withoutthese experiences (NICHD Early Child CareResearch Network, 2002). Positive relationshave also been found between experience informal child care and children's social andbehavioral outcomes (Belsky, 2002; NICHDEarly Child Care Research Network, 2001b).

Although this NICHD Study was de-signed to study the effects of center-based ex-periences for all children including thoseamidst multiple risk factors, it did not intention-ally target children at highest risk in most denseurban settings. For example, children living indangerous neighborhoods in large complex ur-ban settings and those bom to a teen mother wereexcluded from the NICHD study (NICHD EarlyChild Care Research Network, 1994). Therefore,the findings are not necessarily representativeof all children living in our largest urban cen-ters—particularly for children living in the mostdisadvantaged urban neighborhoods.

584


Implications for Future Research andPractice

The current short-term longitudinal,population-based study of center-based childcare and education makes a contribution to theresearch literature by accounting for the rangeof relevant family and neighborhood risk fac-tors while examining the unique relationshipbetween center-based experiences and kinder-garten competency outcomes. This study high-lights the protective benefit of center-basedexperiences for a variety of child outcomes.The objectives of this research were not toevaluate the effects of specific early childhoodprograms, but rather to account for children'sparticipation across the range of early center-based care experiences including federalprograms such as Head Start, private cen-ter-based care, and subsidized child care.Furthermore, this study examined the effectsof these experiences on a broad range ofoutcomes including motor skills, work hab-its, social skills, and attendance.

The present study provides a macro, popu-lation-based assessment ofthe relationships be-tween demographics, family context, neighbor-hood context, early childhood experiences andkindergarten outcomes in a large urban setting.To accomplish this investigation, researcherspartnered with city and school district personnelto capitalize on well-established administrativedatabases. This process enabled large-scale re-search to be conducted within an entire munici-pal system serving young children. However,these databases were limited in the amount ofdemographic information they contained. Futureresearch should incorporate additional parent andfamily variables hypothesized to infiuence earlyschool success, such as parent practices, familyliterary, and parental involvement.

The current study was designed to in-clude a range of center-based early childhoodexperiences, detailed information about spe-cific aspects of the quality of early childhoodprograms was not assessed. Future inquiryshould explore the hypothesized contributionsof individual programmatic features such asthe quality of the learning environment, staffcredentials and training, and curricular foci.Future research efforts could benefit from

employing additional standardized academicachievement and behavioral measures to testmore specific skill sets. Multidimensional as-sessments of early literacy and socioemotionaldevelopment, for example, could provide re-fined information about relationships betweenearly experiences and early school outcomes.In the present study, academic and behavioraloutcome evaluations were based on teacherassessments. In future studies, obtaining infor-mation across multiple informants, such as theparent and outside evaluators, would providemore comprehensive assessments of children'sacademic and behavioral functioning.

The recent No Child Left Behind legis-lation (U, S. Department of Education, 2002)compels early childhood educators and policymakers to focus more attention on youngchildren's transition to elementary school. Evi-dence is needed to document successful pro-grams and processes that foster productivetransitions for the most vulnerable students.Given the keen interest in the impact of HeadStart (U,S. Department of Health and HumanServices, Administration for Children andFamilies, 2005), future studies can examine thespecific effectiveness of Head Start upon en-try to kindergarten and across primary gradeswithin an entire municipal cohort. These quasi-experimental data will add to our understand-ing of Head Start contribution to school readi-ness for children living in poverty.

The current research contributes to thiseffort. It highlights the strategic importance ofpublic school kindergarten in one large urbancenter by relating the unique impact of earlychildhood experiences on specific academicand behavioral competencies for children en-tering school. Furthermore, it underscores theindependent impact of relevant child and fam-ily risks on school readiness competencies.Findings from this study reinforce the impor-tance of center-based early childhood experi-ences to early elementary school success (Na-tional Research Council and Institute of Medi-cine, 2000), Unfortunately, young childrenfrom high-dsk neighborhoods are less likelyto have such experiences (Barnett, 1995). Atthis time, the challenge is twofold. First, thistype of population-based research must be used

585


to foster community-wide discussions aimedat connecting children living in high risk neigh-borhood to center-based early childhood ex-periences. Second, educators, researchers, andfamilies must go beyond the mere recognitionof the importance of formal preschool experi-ence for vulnerable children. Dynamic two-way communications must be generated be-tween formal early childhood programs and thekindergarten programs in the public school sys-tems that receive these children (Pianta, Kraft-Sayre, Rimm-Kaufman, Gercke, & Higgins,2001). Bidirectional relationships provide anopportunity for these early childhood and kin-dergarten programs to share resources and in-formation. This type of collaboration hasgreat promise to improve the quality and con-tinuity of early learning experiences for chil-dren and their subsequent success in school.

Footnotes

'110 Hispanic and 1 Asian student in thissample were receiving services for English Lan-guage Learners.

References

Aureli, T, & Colecchia, N. (1996). Day care experienceand free play behavior in preschool children. Journalof Applied Developmental Psychology, 17, 1-17.

Axinn, W., Duncan, G. J., & Thornton, A. (1997). Theeffects of parents' income, wealth, and attitudes onchildren's completed schooling and self-esteem. In G.J. Duncan & J. Brooks-Gunn (Eds.), Consequences ofgrowing up poor (pp. 518-540). New York: RussellSage Foundation.

Bamett, W. S. (1995). Long-term effects of early child-hood programs on cognitive and school outcomes. TheFuture of Children, 5, 25-50.

Belsky, J. (2002). Quantity counts: Amount of child careand children's socioemotional development. Journalof Developmental & Behavioral Pediatrics, 23, 167-170.

Berry, M. (1992). Contributors to adjustment problems ofadoptees: A review of the longitudinal research. Childand Adolescent Social Work Journal, 9, 525-540.

Broberg, A. G., Wessels, H., Lamb, M. E., & Hwang, C. P.(1997). Effects of day care on the development of cog-nitive abilities in 8-year-olds: A longitudinal study.Developmental Psychology, 33, 62-69.

Brooks-Gunn, J., & Chase-Lansdale, P. L. (1995). Ado-lescent parents. In M. Bomstein (Ed.), Handbook ofparenting (pp. 113-149). Hillsdale, NJ: Erlbaum.

Burchinal, M. R., Campbell, F. A., Bryant, D. M., Wasik,B. H., & Ramey, C. T (1997). Early intervention andmediating processes in cognitive performance of chil-

dren of low-income African American families. ChildDevelopment, 68, 935-954.

Burchinal, M., Lee, M., & Ramey, C. (1989). TVpe of day-care and intellectual development in disadvantagedchildren. Child Development, 60, 128-137.

Burchinal, M. R., Ramey, S. L., Reid, M. K., & Jaccard, J.(1995). Early child care experiences and their associa-tion with family and child characteristics during middlechildhood. Early Childhood Research Quarterly, 10,33-61.

Campbell, J. J., Lamb, M. E., & Hwang, C. P. (2000). Earlychild-care experiences and children's social compe-tence between 1.5 and 15 ytdxsof age. Applied Devel-opmental Science, 4, 166-175.

Campbell, F. A., Pungello, E. P., Miller-Johnson, S.,Burchinal, M., & Ramey, C. T (2001). The develop-ment of cognitive and academic abilities: Growthcurves from an early childhood educational experiment.Developmental Psychology, 37, 231-242.

Campbell, F. A., & Ramey, C. T (1994). Effects of earlyintervention on intellectual and academic achievement:A follow-up study of children from low-income fami-lies. Child Development, 65, 684-698.

Christian, K., Morrison, R J., & Bryant, F. B. (1998). Pre-dicting kindergarten academic skills: Interactionsamong child care, maternal education, and family lit-eracy environments. Early Childhood Research Quar-terly, 13, 50l-52\.

Clarke-Stewart, A. (1991). A home is not a school: Theeffects of child care on children's development. Jour-nal of Sociallssues, 47, 105-123.

Coulton, C , Korbin, J., & Su, M. (1996).Measuring neigh-borhood context for young children in an urban area.American Journal of Community Psychology, 24, 5-32.

Coulton, C, Korbin, J., Su, M., & Chow, J. (1995). Com-munity level factors and child maltreatment rates. ChildDevelopment, 66, 1262-1276.

Coulton, C. J., & Pandey, S. (1992). Geographic concen-trations of poverty and risk to children in urban neigh-borhoods. American Behavioral Scientist, 35,238-257.

Creps, C. L., & Vernon-Feagans, L. (1999). Preschoolers'social behavior in day care links with entering day carein the first year. Journal of Applied DevelopmentalPsychology, 20, 461-479.

Duncan, G. J., Brooks-Gunn, J., & Klebanov, P. K. (1994).Economic deprivation and early childhood develop-ment. Child Development, 65, 296-318.

Egeland, B., Carlson, E., & Sroufe, L. A. (1993). Resil-ience as process. Development and Psychopathology,5,517-528.

Fantuzzo, J., & Cohen, H. L. (2003, April). The EarlyChildhood Experiences Survey: Building capacity toexplore the link between pre-school and kindergartenexperiences. In C. Me Wayne (Chair), Assessment andaccountability in early childhood education: Buildingcapacity through partnership in a large, urban schooldistrict. Symposium conducted at the Annual Meetingof the American Educational Research Association,Chicago, IL.

Farver, J., Natera, L., & Froscsh, D. (1999). Effects ofcommunity violence on inner-city preschoolers and

586


their families. Journal of Applied Developmental Psy-chology, 20. 143-158.

Faul, R, & Erdfelder, F. (1992). GPOWER: A priori, post-compromise power analyses for MS-DOS. Bom, Ger-many: Bom University, Department of Psychology.

Ginshurg, H. P., & Baroody, A. J. (1990). Test of EarlyMathematics Ahility (2-' ed.). Austin, TX: PRO-ED.

Glass, G. Y, Peckham, P. D., & Saunders, J. R. (1972).Consequences of failure to meet assumptions under-lying the fixed effects analyses of variance and cova-riance. Review of Educational Research. 42. 237-288.

High/Scope Educational Research Foundation. (1992).Child Observation Record (COR). Ypsilanti, Ml: High/Scope Press.

Holahan, J. M., Shaywitz, B. A., Chhabra, V., Shneider,A., Marchione, K., Shaywitz, S. E., & Fletcher, J. M.(2(X)2). Developmental trends in teacher perceptionsof student cognitive and behavioral status as measuredby the Multigrade Inventory for Teachers: Evidencefrom a longitudinal study. In D. L. Molfese & V. J.Molfese (Eds.), Developmental variations in teaming:Applications to social executive function, language,and reading skills (pp. 23-52). Mahwah, NJ: LawrenceErlbaum Associates.

Hubbs-Tait, L., Culp, A. M., Huey, E., Culp, R., Starost,H., & Hare, C. (2002). Relation of Head Start atten-dance to children's cognitive and social outcomes:Moderation by family risk. Early Childhood ResearchQuarterly, 17, 539-558.

Innes, F. K., Denton, K. L., & West, J. (2001, April). Childcare factors and kindergarten outcomes: Findings froma national study of children. Presentation at the Societyfor Research in Child Development, Minneapolis, MN.

Klebanov, P K., Brooks-Gunn, J., & McCormick, M. C.(1994). School achievement and failure in very lowbirth weight children. Developmental and BehavioralPediatrics. 15, 248-256.

Kurdek, L. A., & Sinclair, R. J. (2001). Predicting readingand mathematics achievement in fourth-grade childrenfrom kindergarten readiness scores. Journal of Edu-cational Psychology, 93,451-455.

Leventhal, T, & Brooks-Gunn, J. (2000). The neighbor-hoods they live in: The effects of neighborhood resi-dence on child and adolescent outcomes. Psychologi-cal Bulletin. 126, 309-337.

Leventhal, T, & Brooks-Gunn, J. (2003).Children andyouth in neighborhood contexts. Current Directionsin Psychological Science. 12. 273.

Luthar, S. S., Cicchetti, D., & Becker, B. (2000). The con-struct of resilience: A critical evaluation and guide-lines for future work. Child Development, 71,543-562.

McDermott, P A., McWayne, C. M., Fantuzzo, J. W., &Culhane, D. (2003, March). Defining neighborhoodcontext empirically. Paper presented at the EasternPsychological Association's 74th Annual Meeting,Baltimore, MD.

McGraw-Hill. (1997). TerraNova, The Second Editon:Califomia Achievement Tests. Monterey, CA: CTB/McGraw-Hill LLC.

National Association for the Education of Young Children.(1998). Accreditation criteria and procedures of theNational Association for the Education of Young Chil-

dren. Retrieved August 2003 from NAEYC website:www.naeyc.org

National Center for Children in Poverty. (2002). Earlychildhood poverty: A statistical profile. New York:National Center for Children in Poverty, ColumbiaUniversity Mailman School of Public Health.

National Research Council. (2001). Eager to learn: Edu-cating our preschoolers. Washington, DC: NationalAcademy Press.

National Research Council and Institute of Medicine,Board on Children, Youth, and Families, Commissionon Behavioral and Social Sciences and Education.(2002). In J. P Shonkoff & D. Phillips (Eds.), Fwmneurons to neighborhoods: The science of early child-hood development. Washington, DC: National Acad-emy Press.

NICHD Early Child Care Research Network. (1994). Childcare and child development: The NICHD Study ofEarly Child Care. In S. L. Friedman & H. C. Haywood(Eds.), Developmental follow-up: Concepts, domains,and methods (pp. 377-396). New York: AcademicPress.

NICHD Early Child Care Research Network. (2000). Therelation of child care to cognitive and language devel-opment. Child Development. 71, 960-980.

NICHD Early Child Care Research Network. (2001a).Nonmatemal care and family factors in early develop-ment: An overview of the NICHD Study of Early ChildCare. Applied Developmental Psychology, 22.457-492.

NICHD Early Child Care Research Network. (2001b).Child care and children's peer interaction at 24 and 36months: The NICHD Study of Early Child Care. ChildDevelopment. 72, 1478-1500.

NICHD Early Child Care Research Network. (2002). Earlychild care and children's development prior to schoolentry: Results from the NICHD Study of Early ChildCare. American Educational Research Journal. 39,133-164.

Peisner-Feinberg, E. S., Burchinal, M. R., Clifford, R. M.,Culkin, M. L., Howes, C , Kagan, S. L., et al. (2001).The relation of preschool child-care quality tochildren's cognitive and social developmental trajec-tories through second grade. Child Development. 72,1534-1553.

Pianta, R. C , Kraft-Sayre, M., Rimm-Kaufman, S.,Gercke, N., & Higgins, T. (2001). Collaboration inbuilding partnerships between families and schools:The National Center for Early Development andLeaming's Kindergarten Transition Intervention. EarlyChildhood Research Quarterly. 16, 117-132.

The Psychological Corporation. (1990). Differential Abil-ity Scales. San Antonio, TX: Harcourt Brace & Com-pany

Schweinhart, L., McNair, S., Barnes, H., & Larner, M.(1993). Observing young children in action to assesstheir development: The High/Scope Child ObservationRecord study. Educational and Psychological Mea-surement. 53. 445-455.

Schweinhart, L. J., & Weikart, D. P. (1997). The High/Scope Preschool Curriculum Comparison studythrough age 23. Early Childhood Research Quarterly.12, 117-143.

587


Snow, C. E., Bums, S., & Griffin, P. (Eds.). (1998). Pre-venting reading difficulties in young children. Wash-ington, DC: National Academy Press.

Steinberg, L. (1996). Beyond the classroom: Why schoolreform has failed and what parents need to do. NewYork: Simon & Schuster Inc.

Stott, D. H., McDermott, P. A., Green, L. R, & Francis, J.M. (1988). Learning Behaviors Scale and Study ofChildren's Learning Behaviors research editionmanual. San Antonio: The Psychological Corporation.

U.S. Department of Education. (1996). Urban schools:The challenge of location and poverty. Washington,DC: Author.

U.S. Department of Education. (2002). No Child Left Be-hind Act. Washington, DC: Author.

U.S. Department of Health and Human Services, Adminis-tration for Children and Families. (2005, May). HeadStan impact study: Firstyearfindings. Washington, DC.

Weiss, A., & Fantuzzo, J. W. (2001). Multivariate impactof health and caretaking risk factors on the school ad-justment of first graders. Journal of Community Psy-chology, 29, 141-160.

West, J., Denton, K., & Germino Hausken, E., (2000).American's kindergarteners (NCES 2000-070). U.S.Department of Education, NCES. Washington, DC:U.S. Government Printing Office.

Wilson, R. (1987).Risk and resilience in early mental de-velopment. In S. Chess & A. Thomas (Eds.), AnnualProgress in Child Psychiatry and Child Development(pp. 69-85). Philadelphia: Brunner/Mazel, Inc.

John Fantuzzo, PhD, is the Diana Riklis Professor of Psychology in Education, in thePolicy Research, Evaluation atid Measurement Program at the University of Pennsylva-nia. He is also the Director of Penn CHILD Research (Children's Institute of Leaming andDevelopmental Research). Fantuzzo specializes in the population-based study of risk andprotective factors that impact school readiness and early school success for low-incomeurban children.

Heather Rouse, MSEd, is an advanced doctoral student in the School, Community, andClinical Child Psychology program at the Graduate School of Education, University ofPennsylvania. She is an ACF/Head Start Research Scholar, and her research interests in-clude children's socio-emotional development and approaches to learning, kindergartentransition, school readiness, and early childhood public policy.

Paul A. McDermott, PhD, is a Professor in Policy Research, Evaluation, and Measurementat the Graduate School of Education and Professor of Psychology in Psychiatry in theSchool of Medicine at the University of Pennsylvania. He is a psychometrician and quan-titative psychologist. His research interests include psychometrics, test development andscaling, early school emotional and behavioral adjustment, and children's approaches toleaming.

Yumiko Sekino, MA, is an advanced doctoral student in the School, Community, andClinical Child Psychology program at the Graduate School of Education, University ofPennsylvania. She is an ACF/Head Start Research Scholar, and her research interests in-clude early childhood integrative curriculum and assessment, children's peer play skills,and parent involvement in education.

Andrea Weiss, PhD, received her doctorate from University of Pennsylvania in School,Community, and Clinical Child Psychology. She is current serving as a consultant for theOffice of Early Childhood in the School District of Philadelphia. Her current researchinterests are early childhood education and child development.

Stephanie Childs, PhD, received her doctorate from the University of Pennsylvania inhigher education. She is the former director of kindergarten for the Philadelphia PublicSchool District, and current Educational Director of the District's Head Start program.

588

Early Childhood Experiences and Kindergarten Success: A ... › wp-content › uploads › 2015 ›...

Documents

Transcript of Early Childhood Experiences and Kindergarten Success: A ... › wp-content › uploads › 2015 ›...