Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in...

30
This article was downloaded by: [University of North Texas] On: 12 November 2014, At: 11:52 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Early Education and Development Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/heed20 Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty Margaret Tresch Owen a , Julia F. Klausli a , Ana- Maria Mata-Otero a & Margaret O'Brien Caughy b a School of Behavioral and Brain Sciences, The University of Texas , Dallas b Dallas Regional Campus, University of Texas School of Public Health Published online: 23 Apr 2008. To cite this article: Margaret Tresch Owen , Julia F. Klausli , Ana-Maria Mata-Otero & Margaret O'Brien Caughy (2008) Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty, Early Education and Development, 19:2, 302-329, DOI: 10.1080/10409280801964010 To link to this article: http://dx.doi.org/10.1080/10409280801964010 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with

Transcript of Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in...

Page 1: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

This article was downloaded by: [University of North Texas]On: 12 November 2014, At: 11:52Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,UK

Early Education andDevelopmentPublication details, including instructions forauthors and subscription information:http://www.tandfonline.com/loi/heed20

Relationship-Focused ChildCare Practices: Quality ofCare and Child Outcomes forChildren in PovertyMargaret Tresch Owen a , Julia F. Klausli a , Ana-Maria Mata-Otero a & Margaret O'Brien Caughy ba School of Behavioral and Brain Sciences, TheUniversity of Texas , Dallasb Dallas Regional Campus, University of Texas Schoolof Public HealthPublished online: 23 Apr 2008.

To cite this article: Margaret Tresch Owen , Julia F. Klausli , Ana-Maria Mata-Otero &Margaret O'Brien Caughy (2008) Relationship-Focused Child Care Practices: Quality ofCare and Child Outcomes for Children in Poverty, Early Education and Development,19:2, 302-329, DOI: 10.1080/10409280801964010

To link to this article: http://dx.doi.org/10.1080/10409280801964010

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all theinformation (the “Content”) contained in the publications on our platform.However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness,or suitability for any purpose of the Content. Any opinions and viewsexpressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of theContent should not be relied upon and should be independently verified with

Page 2: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

primary sources of information. Taylor and Francis shall not be liable for anylosses, actions, claims, proceedings, demands, costs, expenses, damages,and other liabilities whatsoever or howsoever caused arising directly orindirectly in connection with, in relation to or arising out of the use of theContent.

This article may be used for research, teaching, and private study purposes.Any substantial or systematic reproduction, redistribution, reselling, loan,sub-licensing, systematic supply, or distribution in any form to anyone isexpressly forbidden. Terms & Conditions of access and use can be found athttp://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 3: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

RELATIONSHIP-FOCUSED CHILD CAREOWEN, KLAUSLI, MATA-OTERO, CAUGHY

Relationship-Focused Child CarePractices: Quality of Care and ChildOutcomes for Children in Poverty

Margaret Tresch Owen, Julia F. Klausli,and Ana-Maria Mata-Otero

School of Behavioral and Brain SciencesThe University of Texas at Dallas

Margaret O’Brien CaughyDallas Regional Campus

University of Texas School of Public Health

Research Findings: Child care delivery practices promoting continuous, primarycaregiver–child relationships (relationship-focused child care) were evaluated for223 preschool-age children (45% African American, 55% Latino) attending childcare centers serving low-income children. Both relationship-focused and non-rela-tionship-focused centers were accredited by the National Association for the Educa-tion of Young Children. Children in relationship-focused programs received moresensitive, involved, and affectionate caregiving and were more engaged with theircaregivers than children in comparison centers, but some differences were greater forAfrican American children. Outcomes associated with relationship-focused care in-cluded greater parent-reported child compliance and closer parent–caregiver rela-tions, but no consistent benefits for cognitive school readiness, receptive language, orchild behavior problems were found. Follow-up assessments were completed 1 yearlater for 119 children who remained in their programs. Social and cognitive out-comes improved over time, but some changes were moderated by child race/ethnicityand center type. Over time, parents reported greater child compliance and caregiversreported better parent–caregiver relationships in relationship-focused programs.Practice or Policy: Some social benefits of continuous, primary caregivers were

EARLY EDUCATION AND DEVELOPMENT, 19(2), 302–329Copyright © 2008 Taylor & Francis Group, LLCISSN: 1040-9289 print / 1556-6935 onlineDOI: 10.1080/10409280801964010

Correspondence regarding this article should be addressed to Margaret Tresch Owen, The Univer-sity of Texas at Dallas, 800 West Campbell Rd., GR41, Richardson, TX 75080. E-mail: [email protected]

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 4: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

found, but children’s cognitive competencies improved with sustained attendance atthese accredited programs regardless of the relationship-focused practices.

Based on the premise that children’s experiences in child care are shaped by the re-lationships they hold with their care providers, practices in early childhood pro-grams that promote continuous, positive caregiver–child relationships should en-rich children’s development and well-being. Such practices may be particularlyimportant for children from poor families who are at risk due to less stable life cir-cumstances (Kaiser & Delaney, 1996) and may benefit the children’s adjustment toformal schooling (Birch & Ladd, 1997; Ritchie & Howes, 2003). In this study, weexamined a set of child care program practices that provide a continuous, primarycaregiver–child relationship—practices we have termed relationship-focused. Theeffectiveness of these practices were evaluated for children’s care experiences inchild care and their social and cognitive school readiness in a sample of AfricanAmerican and Latino children who were at risk due to family poverty.

Among the practices that are thought to promote positive relationships betweenchildren and their child care providers, caregiver continuity is considered founda-tional. The goal in providing caregiver continuity is for children to receive carefrom the same caregiver over a prolonged period of time and not experience a se-ries of changing caregivers. When child care center classrooms and their care-givers are designated for a single age, the children change caregivers when theyoutgrow the classroom’s specified age range. Providing care in mixed-age class-rooms where children can remain for a longer period of time is a practice that al-lows greater continuity of caregivers. Designating primary caregivers for the chil-dren in a classroom when there are multiple caregivers is another practice thatshould promote greater continuity in caregiving experiences and positive relation-ships with caregivers. However, primary caregivers do not ensure caregiver conti-nuity. If primary caregivers are assigned but the classrooms are age segregated, achange in primary caregiver will occur when the child is shifted to a new class-room. Similarly, multi-age classrooms in and of themselves will not promote aspecial caregiver–child relationship if the children are not also assigned to primarycaregivers. These different program practices should provide more support for apositive relationship between caregiver and child when they are combined. A spe-cial caregiver–child relationship is more likely to result from practices that allowboth greater continuity of caregivers in mixed-age classrooms and the designationof a primary caregiver for each child.

An emphasis on continuous, primary caregiver–child relationships in child careis based largely on extrapolations from attachment theory. A significant body of re-search relates children’s attachment security with their parents to positive develop-mental outcomes in the areas of psychological adjustment, social competence withadults and peers, self-esteem, and aspects of cognitive and language functioning

RELATIONSHIP-FOCUSED CHILD CARE 303

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 5: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

(e.g., Bus & van IJzendoorn, 1988; Elicker, Englund, & Sroufe, 1992; Greenberg,DeKlyen, Speltz, & Endriga, 1997; Morriset, Barnard, Greenberg, & Booth, 1990;National Institute of Child Health and Human Development [NICHD] Early ChildCare Research Network, 2006b; van IJzendoorn, Dijkstra & Bus, 1995). Another,related, literature stemming from attachment theory has focused on the importanceof the teacher–child relationship in school for children’s positive adaptation toschool and social competencies related to school achievement (see Hamre &Pianta, 2001). It is argued that continuous, primary caregiver–child relationshipsin child care should similarly promote the child’s development.

Supporting this premise, a growing body of research focused on attachment re-lationships between preschool children and their child care providers suggestschildren’s relationships with early childhood program teachers or caregivers con-tribute to their development (see Pianta, 1992, 1999). Children’s attachment secu-rity with nonparental caregivers in child care centers, kibbutzim, and family childcare has been found to be associated with children’s achievement, empathy, andpeer competence (Howes, 1997; Howes, Matheson, & Hamilton, 1994; Op-penheim, Sagi, & Lamb, 1988; van IJzendoorn, Sagi, & Lambermon, 1992). Con-sistent with attachment theory and arguments promoting caregiver continuity, chil-dren who spend more months with their child care provider have been found toexhibit higher interactive involvement and more attachment security with thesecaregivers (Elicker, Fortner-Wood, & Noppe, 1999; Goossens & van IJzendoorn,1990; Howes & Hamilton, 1993; Raikes, 1993). This body of work underlies argu-ments promoting continuity of relationships in children’s child care experience,but there has been little direct study of child care practices that would advancecaregiver continuity or study of associated developmental outcomes for children.

Relatively few child care centers use practices that provide caregiver continuityfor the children (Cryer, Hurwitz, & Wolery, 2000). The prevalent experience forchildren in child care is one of changing caregivers, with a series of different childcare providers and/or multiple caregivers. During infancy and the toddler years,children are moved to new classrooms even more often, with moves coordinatedwith the achievement of certain developmental milestones, such as sitting or walk-ing or becoming toilet trained. Children also change teachers in child care due tohigh rates of teacher turnover (Helburn, 1995; Whitebook, Howes, & Phillips,1998); and some centers move caregivers to different classrooms because atten-dance fluctuates and staffing requirements must be met (Cryer et al., 2000), result-ing in even more changes in caregivers for the children.

Children also experience a lack of caregiver continuity when parents make newchild care arrangements for their children or cobble together multiple arrange-ments to cover nonstandard work schedules and other child care arrangement diffi-culties. Among the children studied in the NICHD Study of Early Child Care andYouth Development (more than 1,150 children studied through their preschoolyears and beyond), the children experienced more than three new arrangements of

304 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 6: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

child care on average across their first 3 years (NICHD Early Child Care ResearchNetwork, 2005), and more than 40% of the infants experienced multiple changes inarrangements in their first year (Tran & Weinraub, 2006). They averaged morethan five different arrangements through 54 months (Belsky et al., 2007). More-over, these averages cannot fully represent the changes children experienced inchild care providers because the measures of child care stability counted the num-bers of new arrangements started but did not include changes within arrange-ments, such as those that are typical in child care centers. In addition, many chil-dren experience multiple, parallel arrangements in the course of the week. DeSchipper, Tavecchio, van IJzendoorn, and van Zeijl (2004) found that a greaternumber of parallel care arrangements interfered with children’s abilities to adaptto the day care setting, particularly when the children had more difficult tempera-ments.

Despite the arguments presented here, empirical evidence for benefits of care-giver continuity with a primary caregiver or risks associated with changing andmultiple caregivers is rare. Two studies, however, are suggestive. Howes and Ham-ilton (1993) found that when toddlers experienced multiple changes in their care-givers in child care, they were more likely to be socially withdrawn or aggressivewith their peers. More recently, Ritchie and Howes (2003) found that spendingmore time with the primary caregiver, together with intense and sensitive care-giver–child interactions, were important predictors of child–caregiver attachmentsecurity. Both studies support the value of caregiver continuity and primary care-givers, but few developmental outcomes have been examined.

The present study examined the effectiveness of program practices promoting acontinuous, primary relationship between children and their caregivers for AfricanAmerican and Latino children living in poverty. The core element of the centers’relationship-focused child care (RFC) practices was grouping the children in“family groups” of mixed ages cared for by a primary caregiver who stayed withthe children for the majority of the day. A family group was composed of a maxi-mum of nine children ranging in age from 3 to 5. Family grouping enabled eachchild to have the same primary caregiver from entry into the center at age 3 untilthe child transitioned to kindergarten, typically 2 to 3 years later. Caregiving prac-tices also included sharing meals, snack times, and activities in the family groups,and individual monitoring and planning for the children by their primary caregiver.Classrooms typically contained two family groups, and larger group activities withthe entire classroom also took place. The RFC centers were Head Start affiliatecenters and enrolled eligible children in the fall following their third birthdays.

A comparison group of children was recruited from centers that enrolled chil-dren with similar sociodemographics but did not include this combination of prac-tices focusing on the caregiver–child relationship. Most of the comparison centershad age-segregated classrooms and children changing classrooms on a yearly ba-sis, but two (of eight) of the comparison centers had multi-age classrooms. The two

RELATIONSHIP-FOCUSED CHILD CARE 305

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 7: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

comparison centers that had multi-age classrooms did not assign primary care-givers. None of the comparison centers employed the family-group practices of theRFC centers. Given our overarching purpose of examining effects of practices pro-moting a continuous, primary caregiver–child relationship, and not confoundingany differences found with overall differences in the quality of the centers, a fur-ther requirement in selecting comparison centers was that all centers be generallycomparable in overall quality. Therefore, both the RFC and the non-RFC centerswere accredited by the National Association for the Education of Young Children(NAEYC), which provided a standard of comparability in quality among the cen-ters (National Center for the Early Childhood Workforce, 1997).

The present study had three goals. The first was to compare observations ofchild care experiences of low-income children attending child care centers withexplicit program practices that enabled and emphasized continuity and closenessin caregiver–child relationships with child care experiences of demographicallysimilar children attending child care centers that did not emphasize continuity inprimary caregiving relationships. We expected that the practices supporting a con-tinuous, close caregiver–child relationship in child care would result in more posi-tive caregiving experiences and greater involvement of the children with their care-givers.

The second goal of the study was to investigate the social and cognitive devel-opment of the children attending RFC and non-RFC centers. We expected that thechildren attending centers emphasizing continuous, positive caregiver–child rela-tionships would have more socially adaptive behavior, fewer behavior problems,and a closer relationship with their child care provider than children attending cen-ters that did not explicitly emphasize continuity in the caregiver–child relation-ship. Because we expected that RFC experience would provide improved supportfor learning, we also predicted that children in RFC would show greater languagecompetence and cognitive school readiness.

The third goal of the study was to examine changes in cognitive and social de-velopment among the children who remained in their centers for at least 1½ years.All children who were initially tested and who remained in their centers 1 yearlater were tested a second time. This longitudinally tested subsample was exam-ined for changes in outcomes over time and for differential changes in relation tothe type of child care they attended—relationship-focused or non-relationship-focused.

METHOD

Participants

A total of 223 three- and four-year-old children (52% boys) were recruited for thestudy from 12 child care centers whose enrollments included 50% to 100% low-in-

306 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 8: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

come children receiving subsidized care. The children had been attending theircenters for at least 6 months prior to recruitment and initial testing. All four of theRFC centers were Head Start affiliate centers, and four of the eight comparisonnon-RFC centers were Head Start or Head Start affiliate centers. All centers oper-ated full-day programs.

The children were first studied in the spring after recruitment (Time 1). Of thechildren studied at Time 1, 123 attended the centers designated as RFC centers and100 children attended non-RFC comparison centers. Of the children attending cen-ters in the RFC group, 43% (n = 53) were African American and 57% (n = 70) wereLatino according to their mothers’ reports. In the non-RFC group, 48% (n = 48)were African American and 52% (n = 52) were Latino. When initially tested atTime 1, children in RFC centers averaged 54.66 (SD = 6.23) months of age and hadattended their centers for an average of 11.69 (SD = 5.22) months; children innon-RFC centers averaged 50.97 (SD = 7.49) months of age and had attended theircenters for an average of 11.42 (SD = 5.14) months. The children were regular par-ticipants in their programs, and only three parents reported using an additionalnonparental care arrangement for their children on a recurring basis. Regular atten-dance was required for continued enrollment in the Head Start programs.

Average monthly income of the families was $1,596.43 (SD = $866.72; annualmean income = $19,157) as reported by mothers. African American children hadsignificantly lower family incomes than the Latino children, F(1, 218) = 4.7, p <.05; children in RFC centers had less family income than children in non-RFC cen-ters, F(1, 218) = 5.92, p < .05. An income-to-needs ratio was determined for eachchild based on the mother’s reported annual family income divided by the govern-ment poverty level for the child’s household size. Income-to-needs ratios averaged1.00 (SD = 0.53). African American children in the RFC group had significantlylower income-to-needs ratios (M = 0.74, SD = 0.41) than children in the other threegroups (Latino RFC, African American non-RFC, and Latino non-RFC). The av-erage number of years of maternal education did not differ (p > .05) between thetwo child care groups (M = 11.67, SD = 2.51, and M = 11.42, SD = 3.05, for theRFC and non-RFC groups, respectively) but was significantly lower (p < .05) forLatino mothers (M = 10.75, SD = 3.13) than for African American mothers (M =12.59, SD = 1.71). More than twice the percentage of the African American chil-dren (65.56%) than Latino children (29.82%) lived in single-parent families.

Approximately half of the recruited children (n = 119; 53.36%) remained intheir centers 1 year later and were assessed a second time (Time 2). Of the childrennot followed at Time 2 (n = 104), 78 had transitioned to kindergarten or a pre-kin-dergarten program and the others had left their centers and were therefore not fol-lowed. Among the sample of children studied longitudinally, 70 attended RFCcenters (48.6% African American, 51.4% Latino) and 49 attended non-RFC cen-ters (34.7% African American, 65.3% Latino). Children who were studied for 2years in their centers did not differ significantly from children who were studied

RELATIONSHIP-FOCUSED CHILD CARE 307

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 9: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

only at Time 1 in terms of their families’ income, their mothers’ education, or childoutcomes assessed at Time 1 (all ps > .05).

Data were also collected from the children’s 61 child care providers. All care-givers had some college education (M = 14.13 years, SD = 1.08) and varied consid-erably in amount of experience in their centers (M = 79.74 months, SD = 83.59)and in age (M = 39.47 years, SD = 11.36). A majority of the caregivers (53%) wereAfrican American, 34% were Latino, and 10% were White. Education, experi-ence, and age of the caregivers did not differ by type of child care.

Procedures

Data collection procedures included the use of individual testing, interviews, ques-tionnaires, and live observations of the children in their child care classrooms.Child outcome data were gathered through individually administered standardizedassessments of receptive language and school readiness, and questionnaires col-lected from caregivers and parents addressing the child’s social behavior andcloseness of the parent–child and caregiver–child relationships. The children weretested individually in their child care centers in areas designated in the centers thatallowed for privacy and few distractions. Family demographic data were collectedin the parent interviews conducted over the telephone. The parent questionnairedata were also collected by telephone interviews to ensure the parents’ under-standing and comfort with the questions and to minimize missing data. Answer op-tions for the questionnaires were described, and each question was read to theparent. Bilingual research assistants conducted interviews in Spanish with the bi-lingual and Spanish-speaking families and tested the children in Spanish, whenappropriate.

Measures

Receptive language. The Peabody Picture Vocabulary Test–Third Edition(PPVT; Dunn & Dunn, 1997) was used to measure children’s receptive languageability. The PPVT is a normed test with good split-half and test–retest reliability.The child is asked to point to a picture named by the interviewer. The items are or-dered from least difficult to most difficult. Split-half reliability for children in theage range of this study is between .77 and .84. Test–retest reliability estimatedfrom a second test administered immediately following the first was between .77and .83. A Spanish version of the test, the TVIP (Test de Vocabulario en ImagenesPeabody; Dunn, Luga, Padilla, & Dunn, 1986), was used with the Spanish/Englishbilingual children. The TVIP was developed from the PPVT–Revised using themost appropriate items for the Spanish population. Norms were developed usingmonolingual students from Latin America, Mexico, and Puerto Rico; the TVIP hasappropriate standardization psychometrics. Given that the children tested in the

308 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 10: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

present study were all bilingual, we altered the test’s standard administration in thefollowing way. Instructions were given to the child in Spanish, and each wordtested was first spoken in Spanish to test the child’s Spanish vocabulary. If the childdid not respond to the word in Spanish, the word was next spoken in English. If thechild identified the correct picture of the word, credit for receptive comprehensionwas given. For the children from English-speaking homes, age-normed standardscores for the PPVT were analyzed; for the children from Spanish-speakinghomes, age-normed standard scores for the TVIP were analyzed.

School readiness. The School Readiness subscale composite of the BrackenBasic Concept Scales (Bracken, 1984) is an age-normed measure of pre-academicskills. It was standardized with a sample of 1,109 female and male children se-lected to be representative of the U.S. population on the basis of sex, ethnic group,geographic region, community size, and socioeconomic status (1980 census). Thecomponents of the School Readiness subscale test the child’s knowledge of colors,letter shapes, number/counting, and comparisons. A standard score is computedbased on the Bracken standardization data. The diagnostic utility of the BrackenSchool Readiness Composite, when used with the Minnesota Preschool Inventory,provided the best prediction of kindergarten performance during the first days ofschool for a sample of 80 preschool children (Zucker & Riordan, 1990). Spanishtest forms and Spanish administration were used in testing the bilingual children.

Behavior problems. The Child Behavior Checklist (Achenbach, 1991a) andthe Teacher’s Report Form (the teacher version of the Child Behavior Checklist;Achenbach, 1991b) were used to measure the children’s internalizing and exter-nalizing behavior problems. Both scales have been standardized on large, repre-sentative samples of children from across the United States; both provide age- andsex-normed standard (T) scores for behavior problems, with cutoff scores for clini-cally significant problems. The scales vary slightly in item content to make themappropriate for the home versus school context. The test–retest reliability, internalconsistency, concurrent and predictive validity, and construct validity of thesescales have been documented in hundreds of studies in North America and in othercultures. Spanish versions of the scale were used with Spanish-speaking parents.The broad band T scores for internalizing and externalizing problems and totalproblems were the examined measures of behavior problems in the present study.

Socially adaptive behavior. The Adaptive Social Behavior Inventory (Ho-gan, Scott, & Bauer, 1992) was used to measure the children’s social skills as re-ported by their mothers and their caregivers. The Adaptive Social Behavior In-ventory is a 30-item questionnaire designed to assess prosocial behaviors inpre-kindergarten-age children. The items are simple and positively worded. Thescale consists of three subscales, two of which assess prosocial behaviors (Express

RELATIONSHIP-FOCUSED CHILD CARE 309

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 11: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

and Comply) and one of which assesses normative/negative behaviors (Disrupt).The standardization sample of 545 children from the Infant Health and BehaviorStudy included 34.9% White, 51.4% Black, and 11.2% Latino low-income chil-dren. The Express and Comply subscales were examined as indicators of sociallyadaptive behavior in the present analyses. Cronbach’s coefficient alphas indicatedacceptable reliability for these subscales and were .69 and .81 for the mother re-ports and .87 and .89 for the caregiver reports for Express and Comply, respec-tively.

Caregiver–child relationship quality. The quality of children’s relationshipswith their caregivers was measured with the shortened Student–Teacher Relation-ship Scale (STRS; Pianta, 1994), a 15-item scale designed to assess teachers’ per-ceptions of a student’s relationship with them. Items were derived from attachmenttheory, the attachment Q-set (Waters & Deane, 1985), and a review of the literatureon teacher–child interactions. A total score is derived from the STRS, indicatingthe overall positiveness of the relationship. In addition, scores for closeness andconflict can be made. In a national sample of 1,022 children, there were no differ-ences on STRS scores related to child ethnicity, but boys scored higher than girlson the conflict scale and lower on the closeness scale (Saft, 1995; Saft & Pianta,2001). The STRS distinguished between high-risk children who had been referredfor special education and those who had not been referred (Pianta, Steinberg, &Rollins, 1995). The total score indicating the overall positiveness of the relation-ship with the caregiver was used in the present analyses.

Quality of child care experience. Observational ratings were made of eachchild’s child care experiences using a coding system developed for the study thatwas adapted from the Observational Record of the Caregiving Environment of theNICHD Study of Early Child Care (NICHD Early Child Care Research Network,2000) and Howe’s Adult Involvement Scale (Howes & Stewart, 1987). Three10-min observation cycles were conducted for each child over the course of at least2 days and included samples of morning and afternoon activities. Each 10-min cy-cle comprised ten 30-s observe/30-s record intervals to obtain frequencies of tar-geted behaviors. Qualitative ratings were made at the end of each 10-min cycle ofobservation.

Each study child was observed separately, and the quality of child care experi-ence was judged from the child’s perspective. Behaviors were rated during each30-s observe segment for each caregiver involved in the child’s care. Frequency be-havioral ratings were averaged across caregivers and across each of the three ob-servation cycles. Analyzed behavior frequency scores represented average fre-quencies of each behavior across a 10-min cycle. Qualitative ratings were alsoaveraged across caregivers and cycles.

310 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 12: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

Caregiver behavior frequencies were determined for caregiver talk to the child;caregiver talk to a group of children, including the target child; caregiver physicalcontact with and physical restriction of the child; affectionate behaviors toward thechild; negativity toward the child; and responses to the child. Child behavior fre-quencies targeted the child’s initiations of positive engagement with the caregiver,neutral/positive interactions with peers, and negative interactions with peers.

Several qualitative ratings were made after each 10-min segment for each care-giver who was in the room with the child at least 5 min during the 10-min segment.Qualitative ratings reflected an overall impression of the quality of the caregiver’s in-teraction with the target child observed during that segment. Qualitative ratings weremade for caregiver sensitivity, cognitive stimulation, detachment, and intrusiveness.

In addition, the level of caregiver involvement was rated in each observation in-terval in which a caregiver interacted with the target child. Involvement ratings of 0indicated that the caregiver ignored the child when the child addressed the care-giver, Level 1 indicated a brief one-word acknowledgement of or response to thechild, Level 2 indicated somewhat more involved and elaborate responses to thechild, and Level 3 ratings included very involved/extended responses from thecaregiver to the child. Frequent Level 0 responses thus indicated a lack of involve-ment with and ignoring of the child; Level 1 responses were indicative of brief,neutral responses; and Levels 2 and 3 indicated progressively more involvedcaregiving. Scores were determined for each level of involvement. The number ofintervals of each level of involvement was based on the average frequency acrossthe caregivers observed per 10-min observation and averaged across the three10-min observation cycles. The scores represented the average frequency of eachinvolvement level the child experienced across a 10-min period.

The child-to-caregiver ratio was observed and recorded just before the start ofeach observation cycle. A score representing the child’s experienced child:care-giver ratio was the average of the three observations of ratio.

All observations were done by four trained observers, all of whom knew aboutthe study’s purpose of comparing experiences and outcomes for children attendingcenters that differed in their program practices regarding caregiver–child relation-ships. To measure interrater reliability, 65 of the 10-min observations were ratedsimultaneously by two raters. Reliability was generally high, with Pearson correla-tions ranging from .79 to .93 for the qualitative ratings and from .67 to .99 for thebehavior ratings. Reliabilities for all but 4 of the 16 qualitative and behavior fre-quency ratings were above .85.

RESULTS

The results are presented in three major sections. The first describes the findingsaddressing differences in the children’s observed child care experiences related to

RELATIONSHIP-FOCUSED CHILD CARE 311

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 13: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

the type of child care program, either relationship-focused or non-relationship-fo-cused. Within this section, comparisons of the process quality measures are de-scribed first, followed by the findings from analyses of child:caregiver ratiosacross the programs and analyses relating ratio to process quality. In the secondmajor section, we present results pertaining to child outcomes and parent–care-giver relationship outcomes measured at Time 1. The final section describes re-sults from analyses of the longitudinally studied subsample, those children who re-mained in their centers 1 year after recruitment and initial assessment.

Comparisons of Observed Child Care Experiencesin RFC and Non-RFC Centers

Quality-of-care observations. The first set of analyses addressed whetherthe care experience differed for children in RFC and non-RFC centers. We ad-dressed this question with a series of 2 × 2 analyses of variance (ANOVAs) testingthe effects of type of child care (RFC vs. non-RFC), child race/ethnicity, and theinteraction of program type and child race/ethnicity on the observation measures.Table 1 reports the mean observation scores by type of program and child race/eth-nicity. Analyses of the qualitative ratings found evidence for higher qualitycaregiving experience in RFC than non-RFC centers as indicated by significantlyhigher qualitative ratings of caregiver sensitivity, F(1, 209) = 29.27, p < .001; andcognitive stimulation, F(1, 209) = 16.55, p < .001. Child race/ethnicity moderatedthe effect of the RFC practices on caregiver sensitivity, F(1, 209) = 6.91, p < .01.Results from a post hoc analysis of means for caregiver sensitivity across the fourgroups indicated that African American children in RFC centers received moresensitive care than African American children in non-RFC centers, but Latino chil-dren’s experiences of sensitivity did not differ significantly by type of center.

Results from the analyses of the behavior frequency ratings similarly indicatedthat the children in RFC centers experienced higher quality caregiving than chil-dren in non-RFC centers according to several indices. Children in RFC centerswere talked to more by their caregivers, F(1, 209) = 8.96, p < .01; and receivedmore affection, F(1, 209) = 9.93, p < .01; more physical restriction, F(1, 209) =5.09, p < .05; and more responses from their caregivers, F(1, 209) = 6.69, p < .05,than children in non-RFC centers. Children in RFC centers also experienced a sig-nificantly higher frequency of Level 3 involvement (the highest level) from theircaregivers (p < .001), but this difference was moderated by child race/ethnicity,F(1, 209) = 4.03, p < .05. Post hoc comparisons of means indicated that high in-volvement occurred more frequently for African American children in RFC cen-ters than for African American children in non-RFC centers but did not differ be-tween Latino children in RFC and non-RFC centers. The frequencies of lowerlevels of involvement did not differ by type of center. Thus, there were multiple in-dications that the children received more positive caregiving in the RFC centers,

312 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 14: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

but two indicators—caregiver sensitivity and high-level involvement—were simi-larly high for Latino children in the RFC and non-RFC centers.

Analyses of the frequency ratings of child behavior found that child negativitydirected at peers was less frequent in RFC than non-RFC centers, F(1, 209) = 5.48,p < .05, but neutral/positive interactions with peers were more frequent for chil-dren in the non-RFC centers, F(1, 209) = 4.39, p < .05. Children in RFC centersinitiated positive engagement with their caregivers more frequently than did chil-dren in non-RFC centers, F(1, 209) = 5.13, p < .05, but this difference was moder-ated by child race/ethnicity, with greater child engagement of caregivers in RFCcenters found only for African American children.

RELATIONSHIP-FOCUSED CHILD CARE 313

TABLE 1Mean Observational Ratings of Child Care Experience by Type of Center

and Child Race/Ethnicity at Time 1

Type of Center

Relationship-Focused Non-Relationship-Focused

Rating LatinoAfrican

American LatinoAfrican

American

QualitativeSensitivitya,b 2.63 (0.44) 2.54 (0.50) 2.44 (0.52) 2.00 (0.51)Intrusiveness 1.07 (0.22) 1.06 (0.21) 0.96 (0.18) 1.02 (0.28)Detachment 1.61 (0.44) 1.60 (0.53) 1.63 (0.64) 1.83 (0.83)Cognitive stimulationa 1.75 (0.57) 1.83 (0.66) 1.55 (0.62) 1.38 (0.42)

Caregiver behavior frequencyTalk to group 10.52 (7.00) 12.61 (9.75) 10.26 (8.48) 10.26 (8.22)Talk to child 5.15 (3.60) 5.98 (4.26) 3.92 (2.98) 4.23 (3.27)Negativity to child 0.12 (0.57) 0.24 (0.74) 0.10 (0.30) 0.40 (0.82)Affection to child 1.40 (1.61) 1.14 (1.42) 0.65 (1.21) 0.70 (1.01)Physical contact with child 0.76 (1.34) 1.00 (1.51) 0.78 (0.95) 1.00 (2.47)Physical restrictiona 0.13 (0.42) 0.35 (1.07) 0.02 (0.14) 0.09 (0.29)Negativity toward child 0.12 (0.57) 0.24 (0.74) 0.10 (0.30) 0.40 (0.82)Response to childa 2.58 (2.20) 3.51 (3.09) 2.33 (2.03) 2.07 (1.84)Involvement Level 0 0.19 (0.47) 0.45 (0.81) 0.20 (0.45) 0.47 (0.83)Involvement Level 1 1.00 (1.36) 1.28 (1.70) 1.14 (1.36) 1.35 (1.40)Involvement Level 2 1.54 (1.65) 1.86 (1.93) 1.35 (1.18) 1.74 (2.01)Involvement Level 3a,b 2.96 (3.03) 3.69 (3.40) 2.39 (2.45) 1.56 (1.91)

Child behavior frequencyNegativity with peersa 0.36 (0.93) 0.82 (1.43) 0.94 (1.48) 1.12 (1.61)Neutral/positive with peersa 12.00 (7.01) 9.33 (6.21) 12.78 (6.77) 12.47 (6.77)Positive engagement with caregivera,b 4.85 (3.87) 5.86 (4.22) 4.88 (4.30) 3.37 (2.94)

Note: N = 212. Standard deviations are in parentheses.aSignificant main effect of type of center (p < .05). bSignificant Type of Center × Child Race/Eth-

nicity interaction effect (p < .05).

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 15: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

Observed child:caregiver ratio. The ratio of children to caregivers ob-served for each child was examined in relation to child care type and child race/eth-nicity in a 2 × 2 ANOVA. There were significant effects for child care type, F(1,211) = 37.34, p < .0001; child race/ethnicity, F(1, 211) = 9.02, p < .01; andrace/ethnicity × child care type, F(1, 211) = 8.90, p < .01. Lower child:caregiver ra-tios were observed for children in the RFC centers, but post hoc analyses of the sig-nificant interaction indicated that only the difference for African American chil-dren was significant (MAfrican American–RFC = 7.24, SD = 2.18; MAfrican American–non-RFC

= 10.53, SD = 3.74; MLatino–RFC = 7.12, SD = 2.22; MLatino–non-RFC = 8.26, SD =2.27).

Ratio was also significantly correlated with several of the observational ratingsof care quality. We therefore examined whether differences in observed quality ofcare attributable to program type remained significant after we controlled for ratiousing a series of analyses of covariance. Despite significant relations to ratio, withobserved features of care indicating higher quality of care associated with lowerchild:caregiver ratios, three observation ratings remained significantly related totype of child care after we controlled for ratio: cognitive stimulation, F(1, 211) =16.21, p < .0001; sensitivity, F(1, 211) = 29.79, p < .0001; and caregiver affection,F(1, 211) = 4.85, p < .05. Even after controlling for ratio, we found that children inRFC centers received more cognitive stimulation, greater sensitivity, and more af-fection from their caregivers, but the differences in sensitivity were again only sig-nificant for the African American children. Involvement Level 3 was nearly signif-icant, F(1, 211) = 3.35, p = .07, and the interaction between type of care andrace/ethnicity was no longer significant (p = .20) after we controlled for ratio.

Child Cognitive and Social Outcomes andParent–Caregiver Relationship—Time 1

Three sets of outcomes measured in the spring of the first year of testing (Time 1)were examined in relation to child care type (RFC vs. non-RFC), child race/ethnic-ity, and the interaction of child care type and child race/ethnicity: (a) individualtests of the children’s cognitive school readiness and receptive language; (b) care-giver and parent reports of children’s social behavior, including behavior prob-lems, socially adaptive behavior, and children’s relationships with the parent/care-giver; and (c) parent and caregiver reports of the parent–caregiver relationship.

Given that child:caregiver ratio differed by child care type and child race/eth-nicity, ratio was a potential confound to relations between child care type and childoutcomes; however, child:caregiver ratio was not significantly correlated with anyof the child outcomes and was therefore not included as a covariate in this set ofanalyses. In addition, we ran analyses of covariance with the same model includingfamily income-to-needs ratio and maternal education as covariates, given that fam-ilies of African American children in non-RFC centers had significantly lower in-

314 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 16: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

comes than other groups and Latino mothers were significantly less educated thanAfrican American mothers. Results of the analyses of covariance did not differfrom those of the ANOVAs, and thus the simpler results from the ANOVAs are re-ported here and in Table 2, with mean scores presented by child care type and childrace/ethnicity.

As shown in Table 2, among the cognitive and language outcomes, significantinteractions between type of center and child race/ethnicity were found for Brackenschool readiness and for children’s receptive language. Child race/ethnicity mod-erated the significant effect of type of center for Bracken school readiness. Posthoc examination of mean scores on the Bracken by child care type and childrace/ethnicity indicated that Latino non-RFC children scored significantly higherthan Latino RFC children, but African American children’s school readinessscores did not differ by type of child care. A similar pattern was found for the chil-dren’s receptive language: Latino children from non-RFC centers scored signifi-cantly higher than Latino children from RFC centers, but African American chil-dren’s receptive language scores did not differ by type of child care. For bothcognitive measures, the Latino children in the non-RFC centers scored higher thanchildren in the other three groups, and there were no indications of benefits of theseRFC practices for the children’s cognitive outcomes.

There was some evidence for differences in parents’ reports of children’s socialbehavior related to attendance at RFC centers. Among measures of social develop-ment as reported by parents, one significant, unmoderated main effect for type ofcenter was found: Children’s socially adaptive compliant behavior as reported byparents was significantly higher in the RFC group (M = 2.51, SD = 0.30) than thenon-RFC group (M = 2.34, SD = 0.38). Other differences related to type of childcare program were moderated by child race/ethnicity. For parents’ reports of chil-dren’s behavior problems, we found significant interactions between type of childcare and child race/ethnicity for both externalizing and internalizing problems.Post hoc analyses of the means, displayed in Table 2, indicated that Latino childrenin RFC centers were rated by their parents to have more internalizing and ex-ternalizing behavior problems than Latino children in non-RFC centers (and thanAfrican American children), but African American children from RFC and non-RFC centers did not differ from one another.

According to analyses of the caregivers’ reports of children’s social behavior,differences were generally inconsistently related or unrelated to type of program.Significant Type of Center × Child Race/Ethnicity interactions were found forcaregivers’ ratings of children’s socially adaptive behavior. Post hoc analyses ofdifferences between the groups indicated that caregivers gave higher ratings forchildren’s socially adaptive behavior on both the Comply scale and the Expressscale for the Latino non-RFC children than for any other group. African Americanchildren did not differ from one another in caregiver ratings of socially adaptivebehavior by type of center. Caregiver ratings of externalizing and internalizing be-

RELATIONSHIP-FOCUSED CHILD CARE 315

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 17: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

316

TAB

LE2

Res

ults

ofA

naly

ses

ofV

aria

nce

and

Mea

nS

core

sfo

rC

ogni

tive

and

Soc

ial–

Em

otio

nalO

utco

mes

byTy

peof

Cen

ter

and

Chi

ldR

ace/

Eth

nici

tyat

Tim

e1

Mea

ns

FL

atin

oA

fric

anA

mer

ican

Out

com

eN

Type

ofC

ente

rR

ace/

Eth

nici

tyTy

pe×

Rac

e/E

thni

city

RF

CN

on-R

FC

RF

CN

on-R

FC

Bra

cken

214

9.39

**0.

845.

13*

88.2

4(1

5.94

) b10

0.04

(15.

87) a

91.2

3(1

5.36

) b93

.00

(17.

12) b

PPV

T/T

VIP

215

1.91

20.1

1***

15.4

0***

90.1

5(1

6.49

) b10

0.82

(15.

85) a

89.0

2(1

1.06

) bc

83.9

1(1

3.43

) cC

BC

L–E

xter

naliz

ing

(par

ent)

197

6.14

*0.

394.

85*

44.5

4(8

.18)

b50

.67

(10.

28) a

46.6

1(9

.35)

b46

.97

(8.0

7)b

CB

CL

–Int

erna

lizin

g(p

aren

t)19

70.

350.

7313

.67*

**47

.62

(9.2

2)b

53.9

8(1

0.53

) a51

.84

(10.

55) b

47.2

2(1

0.97

) bT

RF–

Ext

erna

lizin

g(c

areg

iver

)21

03.

1415

.9**

*2.

9048

.34

(8.5

4)48

.43

(9.9

7)51

.34

(9.6

3)55

.90

(9.5

7)T

RF–

Inte

rnal

izin

g(c

areg

iver

)21

00.

1617

.16*

**3.

8848

.12

(9.2

3)46

.04

(9.3

0)50

.98

(9.7

7)54

.10

(9.3

2)A

SBI–

Exp

ress

scal

e(p

aren

t)19

60.

932.

843.

092.

76(0

.20)

2.68

(0.2

2)2.

76(0

.18)

2.78

(0.2

1)A

SBI–

Com

ply

scal

e(p

aren

t)19

610

.71*

*0.

521.

592.

52(0

.29)

2.30

(0.3

9)2.

50(0

.33)

2.40

(0.3

6)A

SBI–

Exp

ress

scal

e(c

areg

iver

)21

04.

76*

1.33

5.14

*2.

42(0

.39)

b2.

64(0

.24)

a2.

48(0

.33)

b2.

47(0

.39)

b

ASB

I–C

ompl

ysc

ale

(car

egiv

er)

210

1.93

6.8*

*5.

59*

2.39

(0.4

4)b

2.61

(0.3

8)a

2.37

(0.3

9)b

2.32

(0.4

6)b

Car

egiv

er–c

hild

rela

tions

210

0.25

2.95

5.59

*4.

22(0

.62)

ab4.

44(0

.45)

a4.

27(0

.46)

ab4.

13(0

.59)

b

Pare

nt–c

areg

iver

rela

tions

(par

ent)

196

4.06

*4.

46*

2.54

3.62

(0.6

2)3.

58(0

.74)

3.97

(0.6

3)3.

63(0

.61)

Pare

nt–c

areg

iver

rela

tions

(car

egiv

er)

209

1.86

3.33

16.5

6***

3.81

(0.3

9)a

4.06

(0.8

5)b

4.02

(0.6

2)b

3.52

(0.8

0)a

Not

e:Fo

rC

olum

n1,

repo

rter

issh

own

inpa

rent

hese

s.St

anda

rdde

viat

ions

are

inpa

rent

hese

s.D

iffe

renc

esbe

twee

nm

eans

for

type

ofca

reby

child

race

/eth

nici

tyar

esh

own

foro

utco

mes

with

sign

ific

anti

nter

actio

nef

fect

s.M

eans

nots

hari

ngsu

bscr

ipts

diff

ersi

gnif

ican

tly.R

FC=

rela

tions

hip-

focu

sed

child

care

cent

er;P

PVT

=Pe

a-bo

dyPi

ctur

eV

ocab

ular

yTe

st;T

VIP

=Te

stde

Voc

abul

ario

enIm

agen

esPe

abod

y;C

BC

L=

Chi

ldB

ehav

iorC

heck

list;

TR

F=

Teac

her’

sR

epor

tFor

mof

the

CB

CL

;ASB

I=

Ada

ptiv

eSo

cial

Beh

avio

rIn

vent

ory.

*p<

.05.

**p

<.0

1.**

*p<

.001

.

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 18: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

havior did not differ significantly by type of center, but there were significant ef-fects of child race/ethnicity. Caregivers rated African American children higherthan Latino children in both types of behavior problems.

There was a significant Type of Program × Child Race/Ethnicity interaction forcaregivers’ reports of the positiveness of the child’s relationship with them, andpost hoc analyses of the means revealed no significant differences by type of pro-gram for either the Latino or African American children. The most positive care-giver–child relations were reported for Latino children in non-RFC centers and theleast positive were reported for African American children in non-RFC centers.

The third type of outcome examined addressed the relationship between parentsand caregivers, according to parent and caregiver reports. Parents whose childrenattended the RFC centers reported better relationships with their children’s care-givers (M = 3.77, SD = 0.65) than did parents of children attending non-RFC cen-ters (M = 3.60, SD = 0.68). This benefit of RFC was significantly moderated bychild race/ethnicity in the caregivers’ reports of their relationships with the chil-dren’s parents. More positive caregiver–parent relationships were reported in RFCcenters than in non-RFC centers for African American children, but not for Latinochildren. Better parent–caregiver relationships were reported by caregivers of La-tino children in non-RFC centers than in the RFC centers.

Longitudinal Findings

A series of repeated measures ANOVAs was used to examine effects of type ofchild care center (RFC vs. non-RFC), child race/ethnicity, and time of assessment(Time 1 and Time 2) and their interactions for children who were tested a secondtime in the spring 1 year after their initial Time 1 assessments. These analysescould test for within-subject change in child behavior over an approximate year’stime of attendance at the centers; overall type of center and child race/ethnicity ef-fects; and differential change by type of center, by child race/ethnicity, and by typeof center and child race/ethnicity. Approximately half of the children studied atTime 1 had assessments at both Time 1 and Time 2. Mean scores across outcomesby time, type of center, and child race/ethnicity are shown in Table 3. Results of therepeated measures ANOVAs are shown in Table 4.

We found significant improvement in the children’s scores from Time 1 to Time2 for many of the cognitive and social measures: Bracken school readiness (MTime 1

= 92.94, SD = 17.95; MTime 2 = 96.84, SD = 14.55), PPVT receptive language(MTime 1 = 89.90, SD = 15.53; MTime 2 = 94.95, SD = 13.89), parent report of inter-nalizing problem behaviors (MTime 1 = 51.29, SD = 10.38; MTime 2 = 47.77, SD =10.54), caregiver reports of externalizing (MTime 1 = 51.03, SD = 9.93; MTime 2 =48.53, SD = 10.79) and internalizing (MTime 1 = 49.99, SD = 9.95; MTime 2 = 47.26,SD = 10.43) problem behaviors, and caregiver reports of adaptive social behavior

RELATIONSHIP-FOCUSED CHILD CARE 317

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 19: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

318

TAB

LE3

Mea

nO

utco

mes

byTy

peof

Cen

ter,

Chi

ldR

ace/

Eth

nici

ty,a

ndT

ime

Lat

ino

Chi

ldre

nA

fric

anA

mer

ican

Chi

ldre

n

RF

CN

on-R

FC

RF

CN

on-R

FC

Out

com

eTi

me

1Ti

me

2Ti

me

1Ti

me

2Ti

me

1Ti

me

2Ti

me

1Ti

me

2

Bra

cken

88.0

9(1

8.17

)93

.79

(14.

40)

100.

77(1

6.09

)10

0.93

(12.

77)

91.1

5(1

7.14

)95

.82

(15.

55)

91.0

7(1

6.09

)97

.21

(15.

98)

PPV

T/T

VIP

85.3

3(1

2.46

)93

.52

(14.

53)

101.

93(1

6.94

)10

3.63

(11.

35)

86.3

6(9

.86)

91.7

9(1

0.92

)81

.39

(14.

76)

85.3

9(1

3.99

)C

BC

L–E

xter

naliz

ing

(par

ent)

44.5

0(8

.57)

46.2

5(1

0.61

)51

.12

(10.

03)

46.9

2(8

.68)

46.1

3(9

.51)

44.0

4(9

.93)

45.5

0(7

.82)

46.8

0(1

3.07

)C

BC

L–I

nter

naliz

ing

(par

ent)

49.7

5(9

.61)

46.9

6(1

0.71

)54

.96

(9.6

1)49

.28

(9.8

7)52

.63

(10.

54)

48.3

8(1

1.92

)43

.20

(10.

04)

44.8

0(8

.72)

TR

F–E

xter

naliz

ing

(car

egiv

er)

49.6

2(9

.24)

46.0

0(8

.54)

49.5

4(1

0.27

)47

.08

(10.

95)

52.3

2(1

0.44

)52

.36

(12.

23)

58.8

3(6

.43)

49.3

3(1

0.62

)T

RF–

Inte

rnal

izin

g(c

areg

iver

)48

.41

(8.9

9)44

.82

(8.2

5)46

.62

(10.

64)

45.5

8(1

0.92

)52

.97

(9.2

6)51

.16

(11.

65)

58.1

7(8

.75)

48.1

7(8

.86)

ASB

I–E

xpre

ss(p

aren

t)2.

71(0

.22)

2.78

(0.1

7)2.

62(0

.19)

2.77

(0.1

8)2.

76(0

.19)

2.75

(0.1

7)2.

78(0

.15)

2.71

(0.2

5)A

SBI–

Com

ply

(par

ent)

2.47

(0.3

1)2.

57(0

.28)

2.20

(0.4

0)2.

39(0

.38)

2.46

(0.3

5)2.

53(0

.27)

2.32

(0.3

1)2.

22(0

.52)

ASB

I–E

xpre

ss(t

each

er)

2.41

(0.3

7)2.

57(0

.33)

2.69

(0.1

6)2.

65(0

.35)

2.46

(0.3

4)2.

56(0

.36)

2.09

(0.6

0)2.

45(0

.40)

ASB

I–C

ompl

y(c

areg

iver

)2.

36(0

.44)

2.53

(0.3

4)2.

64(0

.40)

2.58

(0.4

3)2.

30(0

.41)

2.42

(0.4

3)2.

12(0

.51)

2.33

(0.5

3)C

areg

iver

–chi

ldre

latio

ns(c

areg

iver

)4.

27(0

.53)

4.40

(0.4

8)4.

58(0

.39)

4.42

(0.6

0)4.

34(0

.47)

4.25

(0.5

5)4.

07(0

.55)

4.05

(0.8

4)Pa

rent

–car

egiv

erre

latio

ns(p

aren

t)3.

87(0

.50)

3.88

(0.6

7)3.

55(0

.76)

3.28

(0.7

1)4.

05(0

.58)

4.06

(0.4

9)3.

62(0

.80)

3.84

(1.3

5)C

areg

iver

–par

entr

elat

ions

(car

egiv

er)

3.73

(0.4

3)3.

91(0

.47)

4.61

(0.7

0)4.

16(0

.26)

3.99

(0.6

0)4.

03(0

.75)

3.71

(1.3

9)3.

61(0

.66)

Not

e:Fo

rC

olum

n1,

repo

rter

issh

own

inpa

rent

hese

s.St

anda

rdde

viat

ions

are

inpa

rent

hese

s.PP

VT

=Pe

abod

yPi

ctur

eV

ocab

ular

yTe

st;T

VIP

=Te

stde

Voc

abul

ario

enIm

agen

esPe

abod

y;C

BC

L=

Chi

ldB

ehav

ior

Che

cklis

t;T

RF

=Te

ache

r’s

Rep

ortF

orm

ofth

eC

BC

L;A

SBI

=A

dapt

ive

Soci

alB

ehav

ior

Inve

ntor

y.

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 20: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

319

TAB

LE4

Res

ults

ofR

epea

ted

Mea

sure

sA

naly

ses

ofLo

ngitu

dina

lOut

com

esin

Rel

atio

nshi

p-F

ocus

edan

dN

on-R

elat

ions

hip-

Foc

used

Pro

gram

s

F

Out

com

eN

Type

ofC

ente

rR

ace/

Eth

nici

tyTy

pe×

Rac

e/E

thni

city

Tim

eTi

me

×Ty

peTi

me

×R

ace/

Eth

nici

tyTi

me

×R

ace/

Eth

nici

ty×

Type

Bra

cken

104

3.10

40.

482.

378.

46**

0.50

0.74

1.49

PPV

T/T

VIP

104

2.59

17.1

8***

15.9

7**

12.0

0***

2.01

0.01

0.82

CB

CL

–Ext

erna

lizin

g(p

aren

t)87

1.27

0.57

0.38

0.84

0.53

0.22

7.02

**C

BC

L–I

nter

naliz

ing

(par

ent)

870.

401.

935.

68*

6.92

*0.

491.

894.

28*

TR

F–E

xter

naliz

ing

(car

egiv

er)

970.

234.

90*

0.07

9.88

**2.

870.

474.

67*

TR

F–In

tern

aliz

ing

(car

egiv

er)

970.

028.

75**

0.15

8.32

**0.

981.

593.

56A

SBI–

Exp

ress

(par

ent)

900.

580.

780.

342.

050.

1211

.19*

*2.

40A

SBI–

Com

ply

(par

ent)

9010

.51*

0.13

0.97

3.11

0.27

4.93

*3.

11A

SBI–

Exp

ress

(car

egiv

er)

990.

196.

88*

8.99

**11

.88*

***

0.10

4.12

*7.

65**

ASB

I–C

ompl

y(c

areg

iver

)99

0.02

8.39

**3.

373.

850.

390.

972.

14C

hild

–car

egiv

erre

latio

nshi

p99

0.13

5.76

*4.

05*

0.25

0.56

0.09

1.77

Pare

nt–c

areg

iver

rela

tions

(par

ent)

887.

17**

2.89

0.21

0.02

0.06

2.95

2.65

Car

egiv

er–p

aren

trel

atio

ns(c

areg

iver

)99

0.87

5.51

*16

.00*

**1.

025.

01*

0.39

2.07

Not

e:Fo

rC

olum

n1,

repo

rter

issh

own

inpa

rent

hese

s.St

anda

rdde

viat

ions

are

inpa

rent

hese

s.PP

VT

=Pe

abod

yPi

ctur

eV

ocab

ular

yTe

st;T

VIP

=Te

stde

Voc

abul

ario

enIm

agen

esPe

abod

y;C

BC

L=

Chi

ldB

ehav

iorC

heck

list;

TR

F=

Teac

her’

sR

epor

tFor

mof

the

CB

CL

;ASB

I=A

dapt

ive

Soci

alB

ehav

iorI

nven

tory

.*p

<.0

5.**

p<

.01.

***p

<.0

01.*

***p

<.0

001.

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 21: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

(Express scale; MTime 1 = 2.47, SD = 0.37; MTime 2 = 2.58, SD = 0.35). Among thesignificant improvements over time for these six outcomes, three were unmodifiedby type of center or child race/ethnicity: the cognitive measures of school readi-ness and receptive language, and caregivers’ reports of child internalizing problembehaviors. In addition, significant Time × Child Race/Ethnicity interactions werefound for parents’ reports of children’s socially adaptive behavior (Express scale)and socially adaptive behavior (Comply scale). Each indicated a significant im-provement in socially adaptive behavior from Time 1 to Time 2 for Latino children(for parent-reported Express scale: MLatino–Time 1 = 2.67, SD = 0.21; MLatino–Time 2 =2.77, SD = 0.17; for parent-reported Comply scale: MLatino–Time 1 = 2.34, SD = 0.38;MLatino–Time 2 = 2.48, SD = 0.34) but no differences over time for African Americanchildren (for parent-reported Express scale: MAfrican American–Time 1 = 2.77, SD =0.18; MAfrican American–Time 2 = 2.74, SD = 0.20; for parent-reported Comply scale:MAfrican American–Time1 = 2.42, SD = 0.34; MAfrican American–Time2 = 2.43, SD = 0.38).

There was only a single significant two-way Time × Type of Center interactionindicating differential improvement by type of center, and this was found for care-givers’ reports of change in parent–caregiver relations over time. Comparisons ofmeans indicated a decline over time in caregiver reports of the closeness of par-ent–caregiver relations in non-RFC centers (MTime 1 = 4.38, SD = 0.99; MTime 2 =4.01, SD = 0.46) but no significant change over time in caregiver reports of thecloseness of parent–caregiver relations in RFC centers (MTime 1 = 3.86, SD = 0.53;MTime 2 = 3.97, SD = 0.62).

Four three-way interactions of Time × Type of Child Care × Child Race/Ethnicitywere found for parents’ reports of children’s externalizing problems and internaliz-ing problems, caregivers’ reports of children’s externalizing behavior problems, andcaregivers’ reports of children’s adaptive social behavior (Express scale). The pat-terning of means for these significant three-way interactions, as shown in Table 3,and results of post hoc analyses did not portray a consistent story. They indicated thata significant decrease in parent-reported externalizing and internalizing problemswas found for Latino non-RFC children and a nonsignificant decrease was found forAfrican American RFC children, but no decrease over time was found for the othergroups. From caregiver reports of children’s social behavior, African Americannon-RFC children showed the greatest improvements over time, with a decrease inreported externalizing problems and an increase in expressive socially adaptive be-havior, but the changes over time were not significant among the other groups.

Between-subject effects of type of center, child race/ethnicity, and type of cen-ter × child race/ethnicity interaction in these longitudinal analyses were similar toresults from analyses of the total, larger sample at Time 1. A significant main effectof center type, unmodified by child race/ethnicity or time, was found for parents’report of the parent–caregiver relationship, indicating closer relations with care-givers for parents in the RFC centers (MRFC = 3.96, SD = 0.57; Mnon-RFC = 3.50, SD= 0.85). Parents also reported higher socially adaptive behavior (Comply scale) for

320 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 22: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

children in RFC centers over both time periods than for children in non-RFC cen-ters (MRFC = 2.52, SD = 0.29; Mnon-RFC = 2.34, SD = 0.39). No other main effectswere found for type of center.

Three significant two-way interactions of Type of Center × Child Race/Ethnicitywere found that were unmodified by time—for the PPVT, the child–caregiver rela-tionship, and caregivers’ reports of parent–caregiver relations. The pattern of differ-ences between groups was similar for these three outcomes and similar to the resultsdescribed above with the larger sample examined at Time 1. The highest scores onthe PPVT over time were found for Latino children in non-RFC centers (M = 102.78,SD = 14.32), and the lowest scores were found for African American children innon-RFC centers (M = 83.39, SD = 14.24). In RFC centers, Latino and AfricanAmerican children’s PPVT scores (MLatino = 89.42, SD = 14.05; MAfrican American =89.07, SD = 10.67) did not differ, but they were significantly below the scores of theLatino children in non-RFC centers. For the child–caregiver relationship overtime, more positive relationships were reported for Latino children in non-RFCcenters (M = 4.50, SD = 0.51) than for African American children in non-RFC cen-ters (M = 4.06, SD = 0.69), but child–caregiver relationships for Latino (M = 4.33,SD = 0.51) and African American (M = 4.30, SD = 0.51) children in RFC centersdid not differ significantly from one another or from the other two groups. A simi-lar pattern was found for caregivers’ reports of parent–caregiver relations. Care-givers of Latino children from non-RFC centers reported more positive relationswith parents (M = 4.39, SD = 0.57) than did caregivers of African American chil-dren from non-RFC centers (M = 3.66, SD = 1.05). Caregivers from RFC centersreported similar degrees of closeness of relations with parents of Latino children(M = 3.82, SD = 0.45) and African American children (M = 4.01, SD = 0.68).

Several main effects of child race/ethnicity, unmodified by time or center type,were found, indicating better social outcomes in Latino than in African Americanchildren according to caregivers. Caregivers reported fewer problem behaviors,both externalizing (MLatino = 48.03, SD = 9.69; MAfrican American = 52.62, SD = 11.00)and internalizing (MLatino = 46.39, SD = 9.60; MAfrican American = 52.24, SD = 10.32);greater socially adaptive behavior according to the Comply scale (MLatino = 2.52,SD = 0.41; MAfrican American = 2.33, SD = 0.44); and more positive child–caregiverrelations for Latino children than for African American children (MLatino = 4.35,SD = 0.55; MAfrican American = 4.21, SD = 0.55). In contrast, African American andLatino children did not differ in their parents’ reports of their social behavior.

DISCUSSION

In this examination of the effects of child care practices designed to provide a con-tinuous, positive relationship between caregiver and child, we found relatively few

RELATIONSHIP-FOCUSED CHILD CARE 321

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 23: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

of the hypothesized benefits for child outcomes in a sample of African Americanand Latino preschoolers from low-income families. For approximately 200 chil-dren from 12 centers that were comparable in terms of their accreditation fromNAEYC, comparisons were made between children attending centers designatedas relationship-focused on the basis of their practices of multi-age family group-ings with a continuous primary caregiver and those designated as non-relation-ship-focused. Only parents’ reports of their children’s socially adaptive compli-ance behaviors were significantly higher for both African American and Latinochildren attending the RFC centers. This effect was found for the children studiedinitially and for the children who were also studied in their centers 1 year later. Nobenefits of the RFC practices were found for the cognitive outcomes of schoolreadiness skills or receptive language, and no consistent benefits of RFC practiceswere found from the caregivers’ reports of the children’s social behavior.

The targeted center practices, however, were associated with stronger par-ent–caregiver partnerships. In the Time 1 assessments, parents reported closer re-lationships with their children’s caregiver when their children attended the RFCcenters—those implementing the practices of family groupings and a continuousprimary caregiver. In the longitudinally studied subsample, caregivers reportedcloser relations with parents across the 2 years of study. The parent–caregiver rela-tionship, marked in part by closer communication between the parent and the care-giver, greater parent involvement, and the parent’s positive evaluation of the child’scaregiver, is considered a marker of higher quality child care (Bredekamp &Copple, 1997). Moreover, when parents and caregivers have closer relations andshare information about the child, parents and caregivers alike have been shown toprovide more sensitive, positive caregiving for the child (Owen, Ware, & Barfoot,2001), enriching care experiences for the child both in child care and at home. Theassociations found with the parent–caregiver relationship indicate an importantbenefit of the centers’ RFC practices.

The observation-based ratings of the quality of child care experience for thechildren indicate that higher quality care was received by children in the RFC cen-ters in terms of caregiver sensitivity and cognitive stimulation, although the differ-ence in caregiver sensitivity was found only for African American children. Interms of specific behavior frequencies, several additional indicators of higherquality care supportive of a nurturing, close relationship with caregivers werefound to be more frequent in the RFC centers. Study children in the RFC programsreceived more talk from their caregivers, more affection, more physical restriction,and more frequent responses than study children in the non-RFC programs. More-over, the children were more positively engaged with their caregivers in the RFCprograms, as evidenced by observations indicating that the children initiated posi-tive interactions with their caregivers more frequently, but this difference, similarto caregiver sensitivity, was found only for African American children. Many ofthese differences in caregiving experiences remained significant after controlling

322 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 24: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

for the effects of child:caregiver ratio, which differed for African American chil-dren in RFC and non-RFC centers. Thus, the RFC program practices of a continu-ous primary caregiver and family-type groupings of the children within the class-room are associated with more positive caregiving experiences for the children.The importance of such experiences has been documented across several studiesshowing that positive caregiving experiences in child care are associated with chil-dren’s social competence and learning (e.g., Hirsh-Pasek & Burchinal, 2006;Howes, Phillipsen, & Peisner-Feinberg, 2000; Howes & Smith, 1995; NICHDEarly Child Care Research Network, 2002) and with secure child–caregiver at-tachments (Ritchie & Howes, 2003).

It is surprising, therefore, that the observed differences between type of centerin the quality of caregiving did not translate into more evidence for beneficial ef-fects of the RFC practices for child outcomes, at least with the concurrent mea-sures of child outcomes examined in the present study. The focus of the observa-tion measure, however, was on the children’s experiences of sensitive, responsivecare, those features of caregiving generally associated with attachment security.The observation rating scales were undifferentiated with respect to the teaching ofspecific skills.

The large number of type of center effects that were modified by race/ethnicitywas unexpected and not easily explained. Controlling for sociodemographic char-acteristics of the families that differed by race/ethnicity did not alter the findings.Several findings indicated benefits of RFC programs for African American chil-dren but benefits of non-RFC centers for Latino children. We looked for explana-tions within the data. The least positive caregiver–child relations were reported forAfrican American children in non-RFC centers, but, contrary to expectations, themost positive caregiver–child relations were reported for Latino children in thenon-RFC centers. In addition, Latino children in non-RFC centers were reportedby their caregivers to be more socially expressive and compliant than Latino chil-dren in RFC centers. Caregivers also reported better relations with Latino parentsin non-RFC than in RFC centers, but better parent–caregiver relations were re-ported by caregivers of African American children in RFC programs. Recall, how-ever, that both African American and Latino parents using RFC centers reportedcloser relationships with caregivers than did parents using non-RFC centers andalso reported greater child compliance.

One conclusion that may be drawn from these disparate findings is that pro-grams and caregivers lacking the kinds of program supports for caregiver continu-ity studied here may nonetheless have other means of supporting positive care-giver–child and caregiver–parent relationships and healthy child development.Moreover, personal characteristics of caregivers and of the children themselvesplay important roles in the quality of care given and received, apart from programfeatures. Consistent with this is the finding that African American children in RFCcenters received greater sensitivity from their caregivers than African American

RELATIONSHIP-FOCUSED CHILD CARE 323

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 25: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

children in non-RFC centers, but caregiver sensitivity did not differ by type of cen-ter for Latino children. A fuller explanation of why RFC program practices are as-sociated with more benefits for African American children and non-RFC practicesare associated with more benefits for Latino children, however, is likely more com-plex and will need further study.

The longitudinal findings obtained for children who remained in their childcare programs in the second year and could be followed in their programs provideevidence for the effects of these child care experiences over a longer period of timefor approximately half of the initially tested sample. The results add to the few ex-isting longitudinal studies of the effects of child care in the community for poorchildren (Loeb, Fuller, Kagan, & Carrol, 2004). In a study of the effects of childcare for low-income children in three communities, Loeb et al. reported positiveeffects for center care compared to home-based care with relatives, and for higherquality and more stable care as defined by changes in arrangements. In the presentstudy of low-income children attending NAEYC-accredited child care centers, wefound that on average the children improved in several respects in both social andcognitive outcomes when they attended their child care programs for an additionalyear, regardless of the type of program they attended (whether relationship-fo-cused or non-relationship-focused).

This is good news for these children, all of whom are at risk for poor schoolachievement due to the very low income of their families and low maternal educa-tion in the Latino families. The improvements found in measures of the children’scognitive and social development over 1 year’s time indicate growth in their readi-ness for school, a primary aim of Head Start and preschool education for childrenfrom low-income families. The cognitive test results for receptive language andschool readiness skills did not show merely growth in the number of words thechildren understood or the school readiness skills they had acquired over a year’stime, which would be expected with age, but improvement in the children’s stan-dard scores on these tests, indicating improvement relative to the standardizednorms.

The children’s growth in vocabulary skills over 1 year in the centers has impor-tant education policy implications because one of the best predictors of reading isvocabulary (National Reading Panel, 2000). These results add to the growing liter-ature on the benefits of child care center experience for children’s language devel-opment. A pertinent example comes from Hart and Risley’s (1995) work tracking42 families for 9 years that showed that one of the best predictors of children’s vo-cabulary was adult talk, including talk in child care settings. From their findings,Hart and Risley concluded that community child care can be an effective interven-tion for low-income children, who typically receive insufficient verbal interactionin their homes.

Similar to existing studies indicating mixed effects of child care quality onsocial–developmental outcomes (NICHD Early Child Care Research Network,

324 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 26: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

2006a), improvements over time in the children’s social behavior across the 2years of measurement were less clear. Significant effects of time on the variousmeasures of social behavior were mostly moderated by child race/ethnicity or byboth race/ethnicity and type of child care, and the direction of change was not con-sistent. For example, the caregivers’ reports of the children’s externalizing behav-ior problems, which declined overall over time, showed the greatest average de-cline for African American children attending non-RFC centers but also declinedbut to a lesser degree for Latino children in both RFC and non-RFC centers. Par-ents’ reports of children’s externalizing problems differed by time of assessment,type of center, and child race/ethnicity, and no consistent pattern over time wasfound across these groups. Improvements in the caregivers’ reports of children’sinternalizing problems over time were not significantly modified by type of centeror child race/ethnicity, but type of center and child race/ethnicity modified the ef-fect of time for parents’ reports. For most groups, caregivers reported fewer inter-nalizing problems for the children at Time 2 than at Time 1, but the largest meandecline was for African American children in non-RFC centers, followed by La-tino children in RFC centers. There was no consistent pattern of relatively greaterimprovement over time in social behavior for children in RFC than non-RFC cen-ters. For socially adaptive behavior, parents of Latino children generally reportedimprovements for their children from Time 1 to Time 2, but there was little changeover time in reports by parents of African American children, and change over timewas unrelated to type of child care program. Thus, changes in social behavior overtime showed various patterns related to type of program, child race/ethnicity, andtime. The inconsistent patterns are likely due in part to the relatively small sizes ofthe groups examined in three-way interactions within the smaller subsample stud-ied longitudinally.

It should be noted that the changes in child outcomes associated with time re-flect changes over 1 year’s time, from the spring of one year to the spring of the fol-lowing year, and therefore do not reflect change attributable to the full extent of thechildren’s child care experience in their programs. Time 1 assessments took placeafter the 3- and 4-year-old children had attended their centers for at least 6 months,but some of the children were in their second year of attendance when assessed ini-tially. Among the children from the four non-Head Start, non-RFC centers, a fewhad been in their centers since infancy.

In sum, in centers serving poor African American and Latino children, childcare program practices designed to provide a continuous primary caregiver in childcare that include family groupings of the children with a primary caregiver andindividualized planning for the children are associated with more responsive andengaged caregiver–child interactions than program practices that are not similarlyrelationship-focused and do not include primary caregivers and mixed-age class-rooms. In addition, we found closer parent–caregiver relations and more sociallyadaptive child compliance for children in the RFC centers, but evidence for other

RELATIONSHIP-FOCUSED CHILD CARE 325

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 27: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

social benefits of the RFC programs was mixed. No benefits of the RFC programpractices were found for the children’s cognitive school readiness or receptive lan-guage skills. Indeed, the children with the highest vocabulary scores were the La-tino children in non-RFC centers. A majority of the Latino children from thenon-RFC centers were from two age-segregated programs that emphasized devel-oping the language skills of the bilingual children they served. Age-segregatedprograms may support a stronger focus on, and better teaching of, language skillsfor bilingual preschool-age children before these children transition to school.

This study adds to the few longitudinal studies of child care with young childrenwho are at risk due to low family income and low maternal education outside of theexperimental studies of particular high-quality care settings. Although we foundrelatively few discernable positive effects for child outcomes related to programpractices supporting a continuous, primary caregiver–child relationship, one shouldnot lose sight of the program differences found in the quality of the children’sday-to-day care experiences and their potential value for these poor children. Lon-ger term follow-up of these children will be important in order to examine whetherany sleeper effects of the children’s preschool child care experiences are discerned,whether the differences found for children’s social behavior and parent–teacher re-lationships continue, and whether different patterns are shown for African Ameri-can and Latino children as the children transition to elementary school.

ACKNOWLEDGMENTS

We wish to thank the participating child care centers, the children, their families,and their caregivers. We also extend our thanks to Connie Aultman, Ginger Brown,Bianca Filio, Michelle Justice, Jose Mendez, Leyna Munoz, Dara Percely, Mar-garita Perez, Fernando Vadillo, and Chunning Wang for their assistance with datacollection and coding. We gratefully acknowledge support from the TimberlawnPsychiatric Research Foundation and ChildCareGroup.

REFERENCES

Achenbach, T. M. (1991a). Manual for the Child Behavior Checklist. Burlington: University of Ver-mont Department of Psychiatry.

Achenbach, T. M. (1991b). Manual for the Teacher’s Report Form and 1991 profile. Burlington: Uni-versity of Vermont Department of Psychiatry.

Belsky, J., Vandell, D. L., Burchinal, M., Clarke-Stewart, K. A., McCartney, K. A., Owen, M. T., andthe NICHD Early Child Care Research Network. (2007). Are there long-term effects of early childcare? Child Development, 78, 681–701.

Birch, S. H., & Ladd, G. W. (1997). The teacher-child relationship and children’s early school adjust-ment. Journal of School Psychology, 35, 61–79.

326 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 28: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

Bracken, B. A. (1984). Bracken Basic Concept Scales. San Antonio, TX: Psychological Corporation.Bredekamp, S., & Copple, C. (Eds.). (1997). Developmentally appropriate practices in early childhood

programs (Rev. ed.). Washington, DC: National Association for the Education of Young Children.Bus, A., & Van IJzendoorn, M. (1988). Mother-child interactions, attachment and emergent literacy: A

cross-sectional study. Child Development, 59, 1262–1272.Cryer, D., Hurwitz, S., & Wolery, M. (2000). Continuity of caregiver for infants and toddlers in cen-

ter-based child care: Report on a survey of center practices. Early Childhood Research Quarterly, 15,497–514.

De Schipper, J. C., Tavecchio, L. W. C., van IJzendoorn, M. H., & van Zeijl, J. (2004). Goodness-of-fitin center day care: Relations of temperament, stability, and quality of care with the child’s adjust-ment. Early Childhood Research Quarterly, 19, 257–272.

Dunn, L. M., & Dunn, L. M. (1997). Peabody Picture Vocabulary Test—third edition. Circle Pines,MN: American Guidance Service.

Dunn, L. M., Luga, D. E., Padilla, E. R., & Dunn, L. M. (1986). Test de Vocabulario en Imagenes Pea-body [Peabody Picture Vocabulary Test]. Circle Pines, MN: American Guidance Service.

Elicker, J., Englund, M., & Sroufe, L. A. (1992). Predicting peer competence and peer relationships inchildhood from early parent-child relationships. In R. Parke & G. Ladd (Eds.), Family–peer relation-ships: Modes of linkage (pp. 77–106). Hillsdale, NJ: Erlbaum.

Elicker, J., Fortner-Wood, C., & Noppe, I. C. (1999). The context of infant attachment in family childcare. Journal of Applied Developmental Psychology, 20, 319–336.

Goossens, F. A., & van IJzendoorn, M. H. (1990). Quality of infants’ attachments to professional care-givers: Relation to infant-parent attachment and day-care characteristics. Child Development, 61,832–837.

Greenberg, M. T., DeKlyen, M., Speltz, M. L., & Endriga, M. C. (1997). The role of attachment pro-cesses in externalizing psychopathology in young children. In L. Atkinson & K. J. Zucker (Eds.), At-tachment and psychopathology (pp. 196–222). New York: Guilford Press.

Hamre, B. K., & Pianta, R. C. (2001). Early teacher–child relationships and the trajectory of children’sschool outcomes through eighth grade. Child Development, 72, 625–638.

Hart, B., & Risley, T. R. (1995). Meaningful differences in the everyday experience of young Americanchildren. Baltimore: Brookes.

Helburn, S. (Ed.). (1995). Cost, quality, and child outcomes in child care centers, technical report. Den-ver: University of Colorado at Denver, Center for Research in Economic and Social Policy, Depart-ment of Economics.

Hirsh-Pasek, K., & Burchinal, M. (2006). Mother and caregiver sensitivity over time: Predicting lan-guage and academic outcomes with variable- and person-centered approaches. Merrill-PalmerQuarterly, 52, 449–485.

Hogan, A. E., Scott, K. G., & Bauer, C. R. (1992). The Adaptive Social Behavior Inventory (ASBI): Anew assessment of social competence in high risk three year olds. Journal of Psychoeducational As-sessment, 10, 230–239.

Howes, C. (1997). Teacher sensitivity, children’s attachment, and play with peers. Early Education andPsychology, 4, 99–107.

Howes, C., & Hamilton, C. (1993). The changing experience of child care: Changes in teachers and inteacher-child relationships and children’s social competence with peers. Early Childhood ResearchQuarterly, 8, 15–32.

Howes, C., Matheson, C. C., & Hamilton, C. E. (1994). Maternal, teacher, and child care history corre-lates of children’s relationships with peers. Child Development, 65, 264–273.

Howes, C., Phillipsen, L. C., & Peisner-Feinberg, E. (2000). The consistency of perceived teacher-childrelationships between preschool and kindergarten. Journal of School Psychology, 38, 113–132.

Howes, C., & Smith, E. W. (1995). Children and their child care caregivers: Profiles of relationships.Social Development, 4, 44–61.

RELATIONSHIP-FOCUSED CHILD CARE 327

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 29: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

Howes, C., & Stewart, P. (1987). Child’s play with adults, peers, and toys. Developmental Psychology,23, 423–430.

Kaiser, A. P., & Delaney, E. M. (1996). The effects of poverty on parenting young children. PeabodyJournal of Education, 71, 66–85.

Loeb, S., Fuller, B., Kagan, S. L., & Carrol, B. (2004). Child care in poor communities: Early learningeffects of type, quality, and stability. Child Development, 75, 47–65.

Morriset, C., Barnard, K., Greenberg, M., & Booth, C. (1990). Environmental influences on early lan-guage development: The context of social risk. Development and Psychopathology, 2, 127–149.

National Center for the Early Childhood Workforce. (1997). NAEYC accreditation as a strategy for im-proving child care quality: An assessment by the National Center for the Early Childhood WorkForce. Washington, DC: Author.

National Reading Panel. (2000). Teaching children to read: An evidence based assessment of the scien-tific research literature on reading and its implications for reading instruction (National Institutes ofHealth Publication No. 004754). Washington, DC: National Institute of Child Health and HumanDevelopment.

NICHD Early Child Care Research Network. (2000). Characteristics and quality of child care for tod-dlers and preschoolers. Applied Developmental Science, 4, 116–135.

NICHD Early Child Care Research Network. (2002). Early child care and children’s development priorto school entry: Results from the NICHD Study of Early Child Care. American Educational Re-search Journal, 39, 133–164.

NICHD Early Child Care Research Network. (2005). Child care and child development: Results fromthe NICHD Study of Early Child Care and Youth Development. New York: Guilford Press.

NICHD Early Child Care Research Network. (2006a). Child-care effect sizes for the NICHD Study ofEarly Child Care and Youth Development. American Psychologist, 61, 99–116.

NICHD Early Child Care Research Network. (2006b). Infant-mother attachment: Risk and protectionin relation to changing maternal caregiving quality over time. Developmental Psychology, 42, 38–58.

Oppenheim, D., Sagi, A., & Lamb, M. E. (1988). Infant-adult attachments on the kibbutz and their rela-tion to socioemotional development 4 years later. Developmental Psychology, 24, 427–233.

Owen, M. T., Ware, A., & Barfoot, B. (2001). Caregiver-mother partnership behavior and the quality ofcaregiver-child and mother-child interactions. Early Childhood Research Quarterly, 15, 413–428.

Pianta, R. C. (1992). Beyond the parent: The role of other adults in children’s lives (Vol. 57). San Fran-cisco: Jossey-Bass.

Pianta, R. C. (1994). Child-Parent Relationship Scale. Charlottesville: University of Virginia.Pianta, R. C. (1999). Enhancing relationships between children and teachers. Washington, DC: Ameri-

can Psychological Association.Pianta, R. C., Steinberg, M., & Rollins, K. (1995). The first two years of school: Teacher-child relationships

and deflections in children’s classroom adjustment. Development and Psychopathology, 7, 295–312.Raikes, H. (1993). Relationship duration in infant care: Time with a high-ability teacher and in-

fant-teacher attachment. Early Childhood Research Quarterly, 8, 309–325.Ritchie, S., & Howes, C. (2003). Program practices, caregiver stability, and child-caregiver relation-

ships. Journal of Applied Developmental Psychology, 24, 497–516.Saft, E. W. (1995). A descriptive study of the Student–Teacher Relationship Scale used with preschool-

ers. Dissertation Abstracts, 55, 3144.Saft, E. W., & Pianta, R. C. (2001). Teachers’perceptions of their relationships with students: Effects of

child age, gender, and ethnicity of teachers and children. School Psychology Quarterly, 16, 125–141.Tran, H., & Weinraub, M. (2006). Child care effects in context: Quality, stability, and multiplicity in

nonmaternal child care arrangements during the first fifteen months of life. Developmental Psychol-ogy, 42, 566–582.

van IJzendoorn, M. H., Dijkstra, J., & Bus, A. (1995). Attachment, intelligence and language: Ameta-analysis. Social Development, 4, 115–128.

328 OWEN, KLAUSLI, MATA-OTERO, CAUGHY

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14

Page 30: Relationship-Focused Child Care Practices: Quality of Care and Child Outcomes for Children in Poverty

van IJzendoorn, M. H., Sagi, A., & Lambermon, M. W. E. (1992). The multiple caretaker paradox: Datafrom Holland and Israel. In R. C. Pianta (Ed.), Beyond the parent: The role of other adults in chil-dren’s lives (pp. 5–24). San Francisco: Jossey-Bass.

Waters, E., & Deane, K. (1985). Defining and assessing individual differences in attachment relation-ships: Q-methodology and the organization of behavior in infancy and early childhood. In I.Bretherton & E. Waters (Eds.), Growing points of attachment theory and research (pp. 41–65). Chi-cago: Monographs for the Society for Research in Child Development.

Whitebook, M., Howes, C., & Phillips, D. (1998). Worthy work, unlivable wages: The National ChildCare Staffing Study, 1988–1997. Washington, DC: Center for the Child Care Workforce.

Zucker, S., & Riordan, J. (1990). Concurrent validity of new and revised conceptual language mea-sures. Psychology in the Schools, 25, 252–256.

RELATIONSHIP-FOCUSED CHILD CARE 329

Dow

nloa

ded

by [

Uni

vers

ity o

f N

orth

Tex

as]

at 1

1:52

12

Nov

embe

r 20

14