Children's Relationships with Caregivers: Mothers and Child Care Teachers Carollee Howes and Claire E. Hamilton University of California, Los Angeles HOWES, CABOLLEE, and HAMU^TON, CLAIKE E. Children's Relationships with Caregivers: Mothers and Child Care Teachers. CHILD DEVELOPMENT, 1992,63,859-866. Infant, toddler, and preschool children's relationships with mothers and teachers were examined in this study. 110 children were observed with their mothers during child care arrivals and reunions. 403 children were observed with their teachers in child care. 3 categories of relationships were derived from these observations. Children in the secure relationship category experienced more teacher involvement than children in the avoidant or ambivalent relationship categories. Children in the ambivalent relationship categories experienced more teacher involvement than children in avoidant relationship categories. A subsample (n = 23) of children were seen with their mothers both during child care arrivals and reunions and in the Strange Situation. Relationship classifications were similar.

The study of caregiver-child relationships in children in child care provides an opportunity to extend our understandings of^ children's development of relationships with adults. Within traditional EuropeanAmerican families, the child's first relationship is to the parent, usually the mother. Within families enrolling their child in child care as an infant, the child simultaneously forms first relationships with parental and alternative caregivers (teachers^). Several recent studies suggest that infiuences on social development may be different in families tiiat use alternative child care for Jieir mfants (Oppenheim, Sagi, & Lamb, 1988; van IJzendoom, Kranenburg, Zwart-Woudstra, & van Busschach, 1990). In tiiese stadies, tiie child s relationship witii tiie teacher as

Much as children's relationships with their parents have a number of interdependent domains or functional aspects (Emde, 1989), children enrolled in child care usually have multidimensional relationships with their teachers. Teachers function as playmates, teachers, managers, and caregivers. j ^ ^ ^ ^ ^^^j^ ^ ^ f^^^^ ^ ^ ^j^^ ^^^^^^^ ^^ ^^^ j ^ ^ ^ ^he qualify of the teacher-child .^i^yonship when the teacher is a caregiver j ^ ^ ^^^.^^^ ^ ^f ^^^ experience of the ^j^.j^ .^ out-of-home child care. In her care^^j ^^ ^^^^^^^ .^ responsible for

T

^^^

u^!,

' ^ r

"^^^ f powerful predictor

known about children's relationships with their teachers. The purpose of this article is to examine these relationships,

^^ ^j^jj^ ^^

^^ ^^^

of children s later social development.

While an extensive and growing lit- diapering, toilet training, and/or handling erature examines mother-child attachment toileting routines; and putting the child to relationships in families using alternative sleep at least once during the day. In a more child care arrangements (see Belsky, 1988; abstract sense, the teacher is responsible for Clarke-Stewart, 1989), relatively littie is keeping the child physically and emotionSome of the teacher-child and mother-child attachment Q-sets were gathered as part of the dissertation of Darlene Galluzzo, the masters theses of Virginia Davila and Ellen Wolpow, and the National Child Care Staffing Study. Thank you for making the data available to us. Thanks also to the research team who worked on the two longitudinal studies: Kristin Droege, Darlene Galluzzo, Annette Groen, Catherine Matheson, Lisabeth Meyers, Jacqueline Moore, Leslie Philipsen, Ellen Wolpow, and Fang Wu. The research could not have been completed without the cooperation of the participating families and child care teachers. Reprint requests may be addressed to Carollee Howes, Graduate School of Education, University of California, Los Angeles, 405 Hilgard Avenue, Los Angeles, GA 90024-1521. ' There is no universal nomenclature to describe the adult who cares for children in child care. She may be a family day-care mother or provider, a caregiver, or a teacher. In this paper we will use the term teacher to mean any alternative to the parent caregiver. [Child Development 1992, 63,859-866. © 1992 by the Society for Research in Child Development, Inc. All rights reserved. 0009-3920/92/6304-0001$01.00]

860

Child Development

ally safe in the absence of the parent. By leaving children with the teacher, the parent explicitly or implicitly tells the child that the teacher is to be his or her main person until the parent returns. Furthermore, children direct proximity maintaining (Goossens & van IJzendoorn, 1990) and social referencing (Camras & Sachs, 1991) behaviors to their child care teachers. This suggests that some children expect child care teachers to provide emotional and physical caregiving. In examining teacher-child caregiving relationships for children enrolled in child care, we chose to use the Waters and Deanne Attachment Q-Set (1985). The Attachment Q-Set (Vaughn & Waters, 1990; Waters & Deanne, 1985) provides for systematic observation of the child's behavior with caregiving adults in ecologically valid settings. The nature of the caregiving relationship between children and adults may change between the infant and preschool periods. For example, Bowlby (1973) suggested that in the preschool years the mother-child relationship becomes a "goalcorrected partnership." We explored the infiuence of the age of the child on the teacher-child relationship in the current study by including a wide age range of children. The Attachment Q-Set was particularly well suited for this purpose, as it is valid throughout the preschool period (Waters & Deanne, 1985). We used Q-Set items to form relationship categories. In order to externally validate these categories, we examined relations between our relationship categories and observations of teacher-child interaction. Two constructs were selected for observation: adult sensitivity and adult involvement. Maternal sensitivity, or the tendency of the mother to respond positively and consistentiy to the infant, is associated with secure mother-child relationships (Ainsworth, Blehar, Waters, & Wall, 1978; Lamb, Thompson, Cardner, & Charnov, 1985). We expected similar relations to hold for teachers in child care. That is, we expected children who appeared emotionally secure with their child care teachers to have more experiences of warmth and positive attention with these teachers than children avoidant or ambivalent with their teachers. Adult involvement may be a particularly useful construct in child care research because each adult in child care programs cares for more than one child. Anderson and colleagues (Anderson, Nagle, Roberts, & Smith,

1981) found that high teacher involvement with children was associated with child behaviors indicative of security in a Strange Situation—like procedure. We validated our maternal relationship categories by comparing our categories with Strange Situation classifications. While observations of child care separations and reunions have been validated with Stiange Situation behaviors (Blanchard & Maine, 1979; Ragozin, 1980), these observations were not categorized with the Q-Set, and standard classifications were not made in the Strange Situation. The Attachment Q-Set has been validated for mothers by comparing QSet assessments made by observers in the home with Strange Situation classifications (Vaughn & Waters, 1990). In tiie current study, we continue the process of measure validation by comparing relationship categories derived from Q-Set observations of mothers as they dropped off and picked up their child at child care with the child's Strange Situation classifications.

Method Sample Four hundred forty-one children participated in this research. One hundred ten children were observed with their mothers, and 403 children were observed with their teachers. Seventy-two of these 403 children were seen with both teacher and mother. This subsample participated in a longitudinal study over a 3-year period, including five data-collection points. Forty-seven children in the longitudinal subsample had data at all points. The children in the full sample ranged in age from 10 to 56 months and were heterogeneous with respect to ethnicity and family structure. The children were all enrolled full time in either child care centers or family day-care homes. Eighty percent of the children entered child care prior to their first birthdays. All of the children had received care from their primary teacher for at least 4 months prior to observation. The quality of child care in the child care arrangements as measured by the Harms and Clifford Environmental Rating Scale (Harms & Clifford, 1980) was extremely varied (M = 3.71, range for total score: 1.10 to 6.90 on a 7-point scale). Procedures Relationships with teacher.—We used the Waters and Deanne (1985) Attachment Q-Set to assess the children's relationship

Howes and Hamilton with their primary teacher. The Q-Set contains 75 items or descriptive statements of the child's behavior toward an adult.^ These items are sorted into nine piles to form a normal distribution of most characteristic (9) to least characteristic (1) descriptive statements. Following pilot work in child care arrangements, we reduced the set of items to 65. The eliminated 10 items were not observed within the child care setting. The validity of the Attachment Q-Set has been established with the Ainsworth Strange Situation (Vaughn & Waters, 1990). In order to complete the Q-Set, two observers each spent at least 2 hours on two separate occasions observing the child with the teacher. We determined the identity of the child's primary teacher by preliminary observations and discussion with the center directors. To obtain children's security scores, the scores from the two observers' Q-Sets were averaged and then correlated with criterion scores for security provided by Waters and Deanne (1985). If the child was between 10 and 35 months old, the 12-month criterion was used. If the child was 36 months or older, the 36-month criterion was used. The correlation coefficients are the children's security scores. A higher score indicates greater security. Interobserver reliability was computed on 60 children not in the study. Children used for reliability were seen simultaneously by observers prior to each data-collection period. Kappa scores on each Q-Set item were computed for interobserver reliability. They ranged from .83 to .95 (median = .92). When two observers saw the same study child at different times, there was somewhat lower agreement between observers. The median kappa score for two observers with one child was .85 (range, .45 to .98). Attachment to mother.—The Attachment Q-Set also was used to assess attachment to mother. Two observers each observed at least two arrivals and reunions as the child was dropped off and picked up at child care. There were large variations among mothers, and also within the same mother on different days, in the length of arrivals and reunions. At the extremes, the longest observation of an arrival or reunion lasted 45 min, and the shortest lasted less than 1 min. Observers completed written de-

861

scriptions of the child during each observed arrival and reunion and did not complete the Q-Set until they felt confident that they had seen sufficient numbers to reliably complete the sort. The median number of observed arrivals was 5 (range, 2 to 10). The median number of observed reunions was also 5 (range, 2 to 11). Interobserver reliability was computed on 30 children. Children used for reliability were seen simultaneously by observers prior to the data-collection period. Kappa scores on each Q-Set item were computed for interobserver reliability. They ranged from .86 to .94 (median = .93). The median kappa score for two observers with one child was .83 (range, .56 to .97). Classification of relationships.—We classified children into three relationship categories. We developed the classification system conceptually by examining correlations between Q-Set ratings and security scores and by cluster analysis of Q-Set items. Both teacher and mother data were used. This process led to profiles of Q-Set items characterizing three types of relationship: emotionally secure, avoidant, or ambivalent. To be classified as secure a child had to receive a 7 or higher rating on all of the following Q-Set items (predominant mood is happy, easily comforted, solicits comfort, greets adults spontaneously, flexible in communication, and obedient), and a 3 or lower rating on all of the following items: unaware of adult changes in location, no physical contact with adult, expects the adult to be unresponsive, and not compliant. To be rated as avoidant a child had to receive a 7 or higher rating on all of the following items: unaware of adult changes, no physical contact, expects adult to be unresponsive, demanding initiation, and a 3 or lower rating on cries often. To be rated as ambivalent a child had to receive a 7 or higher rating on all of the following items: expects adult to be unresponsive, demanding and impatient, distressed social interaction, demanding interaction, and cries often, and a 3 or lower rating on physical contact. Each child was placed in only one category. Strange Situation assessment.—At 12 months, 23 children (11 girls) were seen in a standard Ainsworth Strange Situation procedure. The procedure was completed using

^ Subsequent to our data collection, the Attachment Q-Set has been revised and is now a 90-item sort available from Everett Waters, Psychology Department, State University of New York at Stony Brook.

862

Child Development

the guidelines presented by Ainsworth and associates (Ainsworth et al., 1978). The assessments were videotaped, and the behavior of the infant was coded by observers who had demonstrated reliability with the Minnesota group for Strange Situation assessments. Children were classified as secure (n= 16, 70%), avoidant (n = 3, 13%), ambivalent (n = 1, 4%), or disorganized (n = 3, 13%). Seven months later (mean age of 19.7 months) these children were part of the sample of children observed with their mothers at child care arrivals and reunions. Teacher observations.—An observer, independently from the person who collected the attachment ratings, observed each child and teacher for 2 hours in her classroom. During this period the observer coded four 5-min time samples of the social behaviors of the child. The time samples were spaced evenly throughout the observation period. Each 5-min time sample was broken into 15 20-sec intervals. Within each 20-sec interval the child's proximity to the adult was coded. The child was considered to be in proximity if he or she was within 3 feet of the adult. If the child was in proximity, the adult-child interaction was rated. The adult Involvement Scale (Howes & Stewart, 1987) was used to rate the intensity of the adultchild involvement. This scale has five levels, ranging from (1) routine caregiving in which the teacher touches the child for routine care but makes no verbal responses to the child, through (2) minimal caregiving when the teacher touches or talks to the child in order to discipline the child, to answer direct requests for help, or to give verbal directives with no reply encouraged to more responsive caregiving, (3) answering the child's social bids but not elaborating or extending them, (4) extending and elaborating the child's social bids, and, finally, (5) intense caregiving including holding or hugging the child to provide comfort, engaging the child in prolonged conversation, or playing interactively with the child. For the purposes of this analysis, we used four measures derived from the scale: Percent of low-level involvement (levels 1 and 2); percent of responsive adult involvement (levels 3, 4, and 5); percent of high-level involvement (levels 4 and 5); and percent of intense involvement (level 5). Following the observation period, the observer completed the Arnett Scale of Teacher Sensitivity (Arnett, 1989). The Arnett measure is a 26-item rating scale of teacher sensitivity to children developed for

use in center-based child care. In previous work, the Arnett scale distinguished among teachers with different levels of training in early childhood education (Arnett, 1989). The individual ratings were used to form three composite scores. The composite scores were based on a factor analysis of a larger sample of teachers (Whitebook, Howes, & Phillips, 1990). Sensitive is a composite of nine items, including warm, attentive, and engaged. Harsh is a composite of nine items, including critical, threatens children, and punitive. Scores on the sensitive and harsh scales range from a low of 4 to a high of 36. Detached is a composite of four items, including low levels of interaction, interest, and supervision. Scores range from a low of 4 to a high of 16. Teacher involvement variables were available for 289 children. Arnett teacher sensitivity scores were available for 217 children. Children in the longitudinal subsample had teacher involvement variables for each data-collection point. Interobserver reliability was established to an 82% agreement (agreement/ agreements -I- disagreements) for all behaviors in an interval prior to each datacollection period. Interobserver reliability was reestablished at monthly intervals throughout the entire data-collection period. Median reliability scores from these reliability checks for the adult involvement scale ranged from kappa = .88 to kappa = .94 (median = .92). Median reliability score for the Arnett scale was kappa - .87.

Results Security Scores with Mother and with Teacher The mean security score of children with mothers was .51 (SD = .20, range = - .04 to .78). The mean security score of children with teachers was .30 (SD = .19, range = - .12 to .77). Mother security scores were higher than teacher security scores, t(108) = 8.33, p = .000. Teacher security scores were correlated r = .03 with age. Mother security scores correlated r = .02 with age. There were no sex differences in teacher {t = .49) or mother security scores {t = .52). There was no relation (r = .05) between length of time with teacher and security score. Relationship Categories with Mother and Teacher Seventy-six percent of the children were classified as secure with mother, 14% ambivalent, and 10% avoidant. Seventy-

Howes and Hamilton three percent of the children were classified as secure with teacher, 14% avoidant, and 13% ambivalent. In order to explore differences in relationship categories for different-age children, we compared the age distributions of children in each of our categories. There were no main effects for age in either the teacher, F(2,329) = .41, or mother categories, F(2,108) = .61. A chi-square for association between sex of child and relationship category was not significant. Comparison of teacher behaviors with children of different relationship categories.—We compared teacher involvement

863

and sensitivity with children classified in different teacher relationship categories using multivariate analysis of variance. There was a significant multivariate main effect for classification, F(14,631) = 3.14, p = .001, in the cross-sectional sample (n = 289). Descriptive statistics and univariate F tests for the seven dependent variables are in Table 1. We used Scheffe post hoc tests to compare groups. Children in secure categories were more likely than children in the avoidant cluster and ambivalent categories to have teachers rated lower in harshness and detachment and higher in percent responsive, percent high, and percent intense adult involvement. Children in the teacher avoid-

TABLE 1 COMPAMSON OF TEACHER INVOLVEMENT AND SENSITIVITY AMONG CHILDREN IN DIFFERENT RELATIONSHIP CATEGOBIES RELATIONSHIP CATEGORIES

Cross-sectional sample: Sensitivity: Sensitive Harsh Detach Involvement: Low Responsive High Intense Longitudinal sample: Time 1 (18 month, n = 72): Low Responsive High Intense Time 2 (24 months, n = 63): Low Responsive High Intense Time 3 (30 months, n = 58): Low Responsive High Intense Time 4 (36 months, n = 53): Low Responsive High Intense Time 5 (42 months, n = 47): Low Responsive High Intense * p < .05. ** p < .01.

Secure

Avoidant

Ambivalent

24.2 15.3 6.1

21.4 17.9 7.8

22.4 16.1 6.7

2.14 3.26* 3.63*

12 .17 09 .04

.17 .06 .02 .01

.14 .13 .06 .02

.51 7.53** 4.41** 4.19**

22 .78 72 .17

.37 .63 .43 .07

.37 .72 .67 .09

.85 .88 1.63 3.87**

36 .64 27 .15

.56 .44 .16 .07

.53 .47 .11 .06

2.81 .82 3.59** 5.27**

.14 .86 23 .13

.53 .47 .11 .03

.34 .66 .11 .04

4.39** 5.11** 4.88** 3.59**

27 .72 36 .16

.49 .51 .22 .05

.37

7.80**

29 .71 34 .17

.47 .53 .17 .07

.63

.20 .04

3.66* 3.85**

.39 .61 .22 .10

3.44* 3.45* 3.29* 3.56*

864

Child Development

ant category were more likely than children in the ambivalent category to have teachers rated as harsh and detached. There were significant multivariate main effects for classification at each time period in the longitudinal sample—Time 1: F(8,124) = 2.70, p = .05; Time 2: F(8,110) = 2.62, p = .05; Time 3: F(8,104) = 3.17, p = .003; Time 4: F(8,92) = 2.12, p = .04; Time 5: F(8,92) = 3.15, p = .05. Descriptive statistics and univariate F tests for the four dependent variables are in Table 1. We used Scheffe post hoc tests to compare groups. At Time 1, the children in the teacher secure category received a higher percentage of intense involvement than the children in the teacher avoidant and ambivalent categories. At Time 2, the children in the secure category received higher percentages of highlevel and intense involvement than the children in the avoidant and ambivalent categories. At Times 3, 4, and 5, all levels of involvement were different. The children in the teacher secure category received less Iow-ievel and more responsive, high-level, and intense involvement than children in the other categories. The children in the teacher avoidant category received proportionately less responsive involvement than the children in the teacher ambivalent category. Comparison of Strange Situation classifications for children with different cluster membership.—The subsample of children seen with mothers in the Strange Situation and at child care arrivals and reunions were classified according to Q-Set data as follows: 83% secure, 12% ambivalent, and 4% avoidant. To compare Stiange Situation and Q-Set classifications, we collapsed insecure classifications. Sixty-nine percent (n = 16) of the children received the same maternal classification (secure n = 14, or ambivalent n = 2), using the Strange Situation classification at 12 months and the Q-Set cluster categorization 7 months later (kappa for stability = .49, p < .05). As a validity check on using the Q-Set during separations and reunions at child care, we compared item scores for our 12month Strange Situation secure and insecure groups with those generated by Vaughn and Waters (1990) on a much larger sample of home observations of infants also classified with the Strange Situation. In our procedure, we used 14 of the 16 items that in the Vaughn and Waters (1990) study differentiated secure and insecure groups at the .05

probability level or better. Mean scores and t tests comparing our secure and insecure groups are in Table 2. Seven of the items also differentiated the secure and insecure groups in our sample. The means of remaining items were in the predicted direction, although they did not reach significance. An additional six items differentiated secure and insecure groups in our study. The means for these items also are in Table 2. Discussion The children in the different teacher relationship categories experienced different amounts of teacher sensitivity and involvement. In general, teachers were most sensitive to and most involved with children in the secure category and least sensitive and involved with children in the avoidant category. Children in the ambivalent category experienced less sensitivity and involvement than children in the secure category and more sensitivity and involvement than children in the avoidant category. These differences are consistent with maternal home correlates of Strange Situation classifications (Ainsworth et al., 1978). Research on the development of children in child care settings strongly suggests that children's development is influenced by the quality of child care, and that the quality of care provided is dependent on the education and training of teachers (see Hayes, Palmer, & Zaslow, 1990). In particular, teacher sensitivity and involvement with children are related to their education and specialized training in child development (Whitebook et al., 1990). Future research is needed to further explore relations between teacher characteristics, teacher behaviors, and children's relationships with teachers. For example, do teachers with both secure and insecure children in their classroom behave similarly to these children? If teachers believed that their relationship with the child as well as the child's relationship with the parent was important, would they change their levels of involvement? Children whose attachment to mother was assessed using both the Q-Set—derived cluster membership and the standard Strange Situation tended to receive the same classification as insecure or secure over time across both methods. This offers support for the descriptors we have developed for our categories and suggests that these Q-Setderived categories are based on similar pat-

Howes and Hamilton 865 TABLE 2 ATTACHMENT Q-SET SEPARATION AND REUNION ITEMS DIFFERENTIATING BETvra:EN INFANTS CLASSIFIED SECURE AND INSECURE IN THE STRANGE SITUATION ITEM MEANS Q-SET ITEM

Differentiating in Vaughn and Waters (1990): Predominant mood is happy Transition from exploration to proximity is smooth Affectively responsive Easily comforted Imitates adult Prefers comfort from mother Affective sharing during play Adapts play so does not hurt adult Maintains social interaction Demanding when initiating Lacks self-confidence Expects mother to be unresponsive Object oriented Indirect and hesitant Did not differentiate in Vaughn and Waters (1990): Eager to demonstrate songs, etc Solicits comfort after minor injury Responsive to distress in mother Distressed when social interaction is difBcult Demanding and impatient Cries often

Secure

Insecure

8.38 7.09 7.56 6.57 6.15 5.62 5.47 5.09 4.44 3.84 3.62 2.53 3.25 3.19

6.92 6.50 6.50 4.92 5.25 5.19 4.08 4.58 3.25 5.58 3.83 3.92 3.50 3.25

6.31 5.75 4.97 4.16 3.84 2.13

5.00 4.83 3.33 6.08 5.58 6.00

2.70* 1.71 + 1.74+ 2.64* 1.89+ .70

2.10* 2.62* 2.49* 2.74** .61

3.55** .49 .10

2.53* 2.06* 2.92* 3.01** 2.32* 3.59**

NOTE.—Items ordered by magnitude of secure group's score. + p < .10. * p < .05. **p< .01.

terns of behavior in different settings. This analysis has the same limitation as many other attachment studies. The number of insecure children was too small to permit separate analyses of avoidant and ambivalent children.

distress due to the feelings aroused by child care.

The Attachment Q-Set was designed to be used in home observations (Waters & Deanne, 1985), not specifically to observe sties sful interactions such as separations and Half of the Q-Set items that Vaughn and reunions in child care. The teacher observaWaters (1990) found to differentiate between tions in our study capture the full range of children classified as secure or insecure in daily activities better than the mother obserthe Strange Situation also made the same vations of our study. Future comparisons of differentiation in our sample. This is impor- teacher and mother attachments using the tant because Vaughn and Waters (1990) used Q-Set assessment would do well to use home observations as the basis of their Q- home settings for mothers and child care setsorts. These findings suggest that our Q-sorts tings for teachers in order to confirm our of separations and reunions with motiier at findings. child care tap into the same or similar dimensions as more extensive home observations. Two of the items that differentiated between secure and insecure children in our Referenees study but not in the Vaughn and Waters Ainsworth, M. D., Blehar, M., Waters, E., & Wall, study seem particularly context bound. We S. (1978). Pattems of attachment: A psychomight expect children who wanted to "share logical study of the Strange Situation. Hillstheir day" with their mothers to demonstrate dale, NJ: Erlbaum. songs and other rituals when being reunited Anderson, C. W., Nagle, P., Roberts, M., & Smith, after child care. We also would expect more K. (1981). Attachment in substitute caregivers

866

Child Development

as a function of center quality and caregiver Involvement. Child Development, 52, 53-61. Arnett, J. (1989). Caregivers in day care centers: Does training matter?/owrraoZ of Applied Developmental Psychology, 10, 541-552. Belsky, J. (1988). The "effects" of infant child care reconsidered. Early Childhood Research Quarterly, 3, 235-272. Blanchard, M., & Maine, M. (1979). Avoidance of the attachment figure and social-emotional adjustment in day care infants. Developmental Psychology, 15, 445-446. Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation. New York: Basic. Camras, L. A., & Sachs, V. B. (1991). Social referencing and caregiver expressive behavior in a day care setting. Infant Behavior and Development, 14, 27-36. Clarke-Stewart, A. (1989). Infant day-care: Maligned or malignant? American Psychologist, 44, 266-273. Emde, R. N. (1989). The infant's relationship experience: Developmental and affective aspects. In A. J. Sameroff & R. N. Emde (Eds.), Relationship disturbances in early childhood (pp. 33-51). New York: Basic. Goossens, F. A., & van IJzendoorn, M. (1990). Quality of infants' attachments to professional caregivers: Relation to infant-parent attachment and day-care characteristics. Child Development, 61, 832-837. Harms, T., & Clifford, R. (1980). Early Childhood Environmental Rating Scale. New York: Teachers College Press. Hayes, C. D., Palmer, J. L., & Zaslow, M. J. (1990). Who cares for America's children? Washington, DC: National Academy Press. Howes, C , & Stewart, P. (1987). Child's play with adults, toys, and peers: An examination of family and child care infiuences. Developmental Psychology, 23, 423-430.

Lamb, M. E., Thompson, R. A., Cardner, W., & Charnov, E. L. (1985). Infant-mother attachment: The origins and developmental signflcance of individual differences in Strange Situation behavior. Hillsdale, NJ: Erlbaum. Oppenheim, D., Sagi, A., & Lamb, M. (1988). Infant-adult attachments on the kibbutz and their relation to socioemotional development four years later. Developmental Psychology, 24, 427-433. Ragozin, A. S. (1980). Attachment behavior of day care children: Naturalistic and laboratory observations. Child Development, 51, 409-415. van IJzendoom, M., Kranenburg, M., ZwartWoudstra, H., & van Busschach, A. (1990, April). Day-care and preschool quality of infant-caregiver attachment affects sociability in preschool. Paper presented at the International Society for Infant Studies, Montreal. Vaughn, B. E., & Waters, E. (1990). Attachment behavior at home and in the laboratory: QSort observations and Strange Situation classifications of one-year-olds. Child Development, 61, 1965-1973. Waters, E., & Deanne, K. E. (1985). Defining and assessing individual differences in attachment relationships: 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). Monographs of the Society for Research in Child Development, 50(1, Serial No. 209). Whitebook, M., Howes, C , & Phillips, D. A. (1990). Who cares? Child care teachers and the quality of care in America. Final report of the National Child Care Stafflng Study. Oakland, CA: Child Care Employee Project.

Children's relationships with caregivers: mothers and child care teachers.

Infant, toddler, and preschool children's relationships with mothers and teachers were examined in this study. 110 children were observed with their m...
696KB Sizes 0 Downloads 0 Views