Effects of adult-caregiver's behaviors on the activities, social interactions, and investments of Nascent preschool day-care groups

Effects of adult-caregiver's behaviors on the activities, social interactions, and investments of Nascent preschool day-care groups

JOURNAL OF APPLIED DEVELOPMENTAL PSYCHOLOGY 4, 201-216 (1983) Effects of Adult-Caregiver’s Behaviors on the Activities, SoGal Interactions, and ...

1MB Sizes 1 Downloads 12 Views

JOURNAL

OF APPLIED

DEVELOPMENTAL

PSYCHOLOGY

4,

201-216 (1983)

Effects of Adult-Caregiver’s Behaviors on the Activities, SoGal Interactions, and Investments of Nascent Preschool Day-Care Groups* ANGELA

TZELEPIS, PAUL T. GIBLIN, AND SAMUEL J. AGRONOW Wayne State University

Changes in 16 preschool children’s activities, social interactions, and degrees of initiative and investment are reported in the first and fourth weeks of shared day care experience. The effeck of adult coretakers’ involvements with these children were compored in two day care centers (eight children from each center). Seven adult caregivers from these centers (A and B) were observed to assess their interactions with the children. Adults in center B disployed a significantly greater number of adultinitiated contacts of a more positive nature than did center A adults. Univariate ANOVA’s of children’s behoviors reported significant time, center, ond center-bytime differences. Between weeks 1 and 4 children displayed an increase in simultoneous involvement with peers, adults, and activities; an increase in the number of contacts with peers and adults; a decrease in time spent in transition between activities; and an increase in investment. The number of peer contacts decreased from weeks 1 to 4 in the center with fewer adult-initiated contacts and increased in the high adult-contact center. Descriptions of affiliated network and a sequentiol analysis of state transitions further illustrated the effects of adult behavior on peer contacts and activity selections. Implications of the results are discussed for day care procedures and policies.

An ecological transition occurs when an individual enters a new setting, assumes a new role, or both. Bronfenbrenner (1979) describes an ecological transition as being “a ready-made experiment of nature with a built-in, before-after design in which each subject can serve as his own control” (p. 27). In recent years a few studies have investigated the ecological transition of a newcomer’s entry into a preestablished nursery school (McGrew, 1972; Feldbaum. Christenson, & O’Neal, 1980). In contrast, the present study reports changes in the behavior of two newly formed preschool groups (i.e., all children are “newcomers”) in their first month * Thispaperis based on a thesis completed by the first author in partial fulfillment of the requirements for thedegree of Masterof Artsat Wayne State University. The second author directed this thesis. Appreciation is expressed to Karen Lulewicz, who assisted in data coding; Ira Firestone and Douglas Snyder, who critiqued an earlier version of this paper; and to Leanne Poffenberger who provided insights into the applied implications of the results. Correspondence and requests for reprints should be sent to Angela Tzelepis, Psychology Department, Wayne State University, Detroit, MI 48201.

201

202

TZELEPIS,

GIBLIN, AND AGRONOW

of shared day care experience. We report this transition in two day care centers which varied in the program orientations of the head teachers. Children’s behaviors observed in the present study characterize what children do, with whom they do it, and how they go about it. That is, respectively, behaviors representative of activity selection and performance, social participation, and initiative and investment are observed. Additionally, the co-occurrence of activity involvement and social interactions are recorded. These behaviors were suggested by previous studies of a newcomer’s entry into an established nursery school. McGrew (1972) reports that in the first five days in the nursery social interactions steadily increased, while nervous exploration steadily decreased. That is, the number of activities and the number of children with whom the newcomer interacted increased. Uncertainty in the selection of an activity may be assessed by a comparison of time spent in transition between activities and the frequency and duration of the activities in which the children engage (Stodolsky, 1974). The simultaneous increase of both activity involvement and social contacts imply that acclimation to a new setting may also entail the increased co-occurrence of these two domains. A number of studies have reported the effects of adult day care provider’s behaviors on preschool children’s behaviors (Anderson, Nagle, Roberts, & Smith, 1981; Beller, 1969; Fagot, 1973; Huston-Stein, Friedrich-Cofer, & Susman, 1977; Thompson, 1944). These studies report that caregiver’s degrees of directedness, displays of warmth, uses of praise or criticism, varying of activity, and interactions with children affected a wide range of children’s behaviors including task performance, exploratory behavior, and the establishment of social contacts with both adults and peers. Adult caregiver’s interactions with children are likely to vary with program orientation (Berk, 1971; Miller & Dyer, 1975). The program orientation of the two day care center’s head teachers varied in the degree and nature of adult-child interactions, and the parameters for children’s use of materials. We expected that these differences in orientation’s would affect the above child behaviors. In the present study, changes in children’s behaviors in their first month of shared day care experiences are assessed using measures of frequency and duration (e.g., activity, social contact, etc.), descriptions of social contacts by the construction of affiliative networks (Strayer, 1980), and the prediction of activities from preceeding events by the sequential analysis of state transitions (Gottman & Bakeman, 1979). Additionally, frequency and nature of adult-child interactions are assessed, and their relationships to children’s behaviors are described. METHOD Subjects

Sixteen preschool children were observed in their first and fourth weeks (time 1 and time 2) of shared day care experience in two child care centers located in a large urban community medical complex. In the first week of operation, center A had eight children (five girls and three boys) between the ages of 38 to 60 months (% = 48.5); center B also had eight children (four girls and four boys) between the ages of

SOCIAL INTERACTION

203

29 to 50 months (x = 38.1). Both centers continued to enroll children during the weeks between observations. After four weeks of operation center A had 14 children (ii = 42.9) and center B had 9 children (% = 39.6). However, only the 16 children that were present for both times were included in the present longitudinal analysis. As determined by the head-of-the-households’ occupation, 11 of the 16 children’s families were middle class (e.g., college educated, professional occupations) and 5 were lower class (e.g., supported by a social assistance program). Eight of the families had both mother and father living at home and eight had only the mother living at home. No significant differences existed between centers on these variables. Prior nursery school experience ranged from 0 to 54 months (3 = 9.38). No significant difference existed between centers on this variable nor was it significantly correlated with age. The seven observed adults (three from center A and four from center B) were six females and one male ranging in age from their middle twenties to early fifties. Five were white and two were black. The two head teachers had masters’ degrees in child development; the five staff personnel were enrolled in graduate child development programs. Day Care Centers Centers A and B were similar in size, child-adult ratio (3.5:1), and materials available for play. The major portion of both centers was a 20’ X 40’ room in which the majority of equipment and activities were located. Observations of children in the present study were limited to this room. Both centers were administered by the same university department. Both centers were equipped with similar materials including tables, chairs, floor pads, climbing apparatus, play kitchen equipment, and reading material. Space was provided for easy movement between these materials. Activities varied daily and included cut-and-paste tasks, lego, music, and water play. Children had unlimited opportunities for access to these materials during free play periods. Daily, adults arranged materials for children’s activities. Children were encouraged to select their own activities but adults would aid children who found selection difficult. The primary goal of these day care programs was to facilitate children’s development of a sense of their own needs and the ability to express their wants. Concurrently, children were encouraged to perceive the needs of other children. To these ends, adult-child and peer interactions were encouraged throughout the day. Center policies were promulgated in staff orientation programs and in the instructional material of courses utilizing these centers. As previously noted, the head teachers of these centers were trained in differing program orientations. One had been trained by the department which administered these centers; the other had been trained in an orientation which allowed for limited interactions between adults and children, and restricted the use of materials to predetermined locations and activities. While the latter head teacher initially expressed agreement with the prevailing policies of the centers, preliminary observations indicated her behaviors, and the behaviors of her staff, were at variance with these policies. The present study sought to quantify those differences and access

TZELEPIS, GIBLIN, AND AGRONOW

204

their effects on children’s behaviors during their first month of shared day care experience. Three aspects of this sample should be noted. Firstly, the first week of shared day care experience followed an orientation week in which, on each day, two or three children and their parents attended the day care centers to familiarize tbemselves with the setting and the staff. These sessions generally lasted a few hours. Secondly, while there is significant difference in the ages of the eight children observed in each center, the ages of the total center populations are not significantly different. Therefore, the adult caregivers would presumably behave similarly toward their total groups in terms of their general level of interactions and the form that these interactions assumed. However, effects of age will be assessed for the children’s behaviors. Thirdly, the sample is small and caution should be used in generalizing these results. However, with intensive naturalistic studies it is not normally feasible to have a sample substantially larger than ours. For example, in the two prior studies of newcomers in a nursery school, McGrew (1972) and Feldbaum et al. (1982) had newcomer samples of 12. Procedure Thirty minutes of morning free play activity for each of two days in the first and fourth weeks of shared day care experience were videotaped. A wide-angle lens placed in an observation room adjacent to the main rooms of the day care centers afforded a simultaneous recording of all children and their activities. The location of the observation rooms in both centers was at the end of the 40-foot rooms. For both centers, this placement afforded equal benefits and liabilities in behavioral encoding from videotape. Two limitations characterized this procedure. First, although sound in the centers was recorded it was sometimes difficult to isolate the source of a specific conversation or to hear the content of conversations on the far side of the room. Therefore, behavioral coding relied heavily on visually observable behavior. Second, bathrooms and a small adjoining room in both centers were not in the camera’s range. Procedures used to deal with these limitations are discussed in subsequent behavioral encoding sections. To avoid the children’s reaction to the novelty of a camera (the lens could be seen by the children through a window of the observation rooms), videotaping equipment was placed in the centers several days prior to actual filming. From our observations during times 1 and 2, children paid little if any attention to the presence of this equipment. Behavioral

Encoding

Child Coding. One hundred twenty minutes of videotaped observations were viewed twice for each child. (Due to varying patterns of attendance, 12 children were present all four days [ 120 minutes], five were present on three days, and one child was present for only one day at both times 1 and 2 [60 minutes].) The first viewing observed and recorded behavior in two-minute segments. The second viewing scored behavior by a continuous record using an event recorder. The nature and

Behavior

TABLE 1 of Child Dependent

Definitions

Variable

Variables

Coding Scheme

Definition

Observer Reliability Coefficient

Activity Selection and Performance Transition

Percent of time observed

between ac-

Continuous

.95b

Two-minute

.97a

Continuous,

.92b

tivities (total time less time spent in activity). Average

Area

Movement

Average

number of movements be-

tween activity areos of the centers.

Social Interactions Average

Peer

Duration Frequency of Peer Contacts Average Adult Duration Frequency of Adult Contacts Specific Peer and Adult Contacts

Time observed with peers divided by

Two-minute

number of peer contacts. Average

occurrence of peer contacts

per two-minute

.97=

Continuous,

.90b

segment.

Time observed with adults divided by number of adult contacts. Average

.84a

Two-minute

Two-minute

.97c

Two-minute

.91a

Two-minute

.94a

Two-minute

.79a

activities or so-

Two-minute

,820

in activity divided by

Continuous,

.95b

occurrence of adult contacts

per two-minute

segment.

Names of each peer and adult contacted during each two-minute

seg-

ment. Used to construct affiliative networks.

Initiative Initiative

and Investment

initiation versus following

of activity

or social interaction. Level of

Attention

Investment Average

Activity

Duration

to ongoing

cial interactions. Time observed

number of activities engaged in.

Co-occurrence Peer-Adult

Two-minute

.68c

of Behavion

Percent of time observed with peers

Continuous

.92b

Continuous

.92b

and adults simultaneous. Peer-Activity

Percent of time observed with peers while engaged in activity.

Adult-Activity

.95b

Percent of time observed with adults

Continuous

.9Ob

Continuous

.92b .90b

while engaged in activity. Peer-AdultActivity

Percent of time observed with peers ond adults simultaneously gaged in activity.

while en-

.95b

Olnterrater reliability level. bSpearman-Brown prediction formula. CPeorson Product Moment correlation.

205

206

TZELEPIS,

GIBLIN, AND AGRONOW

range of behaviors observed in this study required these two separate procedures. Table 1 summarizes these behaviors. Five points clarify Table 1. (a) Activity occurred when a child was either actively engaged in a task or play, or was paying attention to an adult’s instructions. Activity was coded as a single episode when a child was engaged in an activity and diverted attention from it for less than one minute. If attention was diverted for more than one minute with the subsequent return to the previous activity a new episode of activity was recorded. Duration of activity was reduced by such fluctuations in attention. (b) A daily inventory was taken of the number, type, and location of available activities for each center. The number of child movements between these activity areas was recorded. Generally, each center was characterized by 8 or 9 predominant activities. (c) Peer or adult interactions occurred when the child initiated or was the recipient of a social contact (e.g., speech, touch) or when the child maintained proximity (within five feet) and oriented towards a peer or adult for a 20-second period. (d) Initiation occurred when a child selected an activity which was not ongoing (i.e., no other child was involved in the activity at the time of selection), a child evolved an ongoing behavior into a new application or form (e.g., the child paints a second picture), or a child initiated an interaction with a peer or adult. (e) Following occurred when a child selected an activity which was ongoing (i.e., at least one other child was already involved with this activity at the time of selection), or a child followed another’s directions or responded to another’s overtures for social contact. No score was given if the child continued with an activity which began in a prior two-minute segment. Adult Coding. A separate coding of adult-child interactions -was performed. Each adult was viewed for 30 minutes: 15 minutes from time 1 and 15 minutes from time 2. To insure that each adult was observed for a total of 30 minutes, coding was discontinued when the adult was “off-camera.” Adult-child interactions were coded in three categories: (a) Initiation of a social contact was inferred from an overt gesture (e.g., touching, reaching, talking). The presence of a response to another’s overture for social contact was determined by an overt gesture of recognition, changed attention, or an alteration of ongoing behavior. (b) Appropriateness of an adult’s response to a child’s initiative involved the prompt response (5 to 10 seconds) to a request for help or social contact. (c) The specific form of an adult’s initiated contact was summarized into two groupings: restricting vs. enhancing contact. When disciplining a child, an adult restricts the child’s behavior. In contrast, teaching, praising, comforting, greeting, and giving affection are behaviors which enhance the child’s feelings of self-worth and may extend the range of their behavioral expression. The variable “restricting behavior” was created to summarize the form of adult-initiated behavior by dividing the number of episodes of disciplining by the number of episodes of the remaining above behaviors. Table 2 summarized the behavioral categories by which adults were observed. Reliabilities For children and adults, 10% of the two-minute and continuous record codings were simultaneously scored by two observers. Three types of reliability coefficients were

SOCIAL

Behavioral

Definitions

207

INTERACTION

TABLE 2 of Adult Dependent

Variables Observer Reliability Coefficienta

Definition

Category Initiator

Person

Response

Response

initiating

interaction

.91

(adult or child).

of recipient of contact (elaborates

continues contact or terminates

ond

.94

further

contact). Appropriatenessb

Appropriateness

of adult response to child-initi-

.82

ated contact. If adult initiated contact, the behavior

was characterized

Task relevant

Teaching

Discipline

Correcting

a child’s behavior.

Praise

Expressing

approval

by one of the following

.93

or directing a child. for something

categories:

.95 a child has

.85

said or done. consolation

.82

Comfort

Giving

Greeting

Displaying

readily perceived greeting behavior.

or encouragement

.89

Affection

Expressing

affection.

.87

alnterrater reliobility level. bAn inappropriate response is when an adult failed to respond contact from a child, or an adult criticizes o child.

to o request for help or social

calculated. Interrater reliablity coefficients, figured as the number of agreements/agreements plus disagreements, were calculated for variables of the twominute and event coding (Billman & McDevitt, 1980). Interrater reliability ranged between .79 for initiative and .90 for specific activity. The reliability coefficients for the duration variables of the continuous record coding ranged between .90 for adult duration and .99 for off-camera duration and were calculated by the Spearman-Brown prediction formula (Winer, 1971). The reliability coefficients for the frequency variables range between .68 for activity frequency and .97 for off-camera frequency and were calculated by the Pearson-Product Moment correlation. Tables 1 and 2 report these reliabilities for each category of observation. RESULTS

The results are presented with three major foci: (a) combined center analysis; (b) center differences analysis; and (c) additional measures of social interaction. The combined center analysis examined changes in behavior between times 1 and 2 without regard for center effects. Change was also analyzed for sex and ages effects. The center differences analysis examined questions regarding differences in behavior of adults in the two centers and the effect of center on changes in children’s behavior. Additional measures of social interaction (i.e., affiliated networks and sequential analysis of state transitions) elaborate and clarify results described in the first two analyses.

TZELEPIS,

208 Combined

Center Analysis:

GIBLIN,

AND

AGRONOW

Time Effects

Univariate mixed-design ANOVAs to assess the effects of age and sex failed to report any significant main effects of these variables or interactions effects of ageby-time or sex-by-time. Table 3 reports a repeated two-by-two univariate mixed design analysis of variance for day-care center-by-time. Significant main effects for time were reTABLE Univariate

Analysis

3

of Variance: Group

Center

A

(n = 8)

Care Center-by-Time

Day

Means

F

Center B (n = 8)

Center

Time Variables

7

Time 2

Time 7

Time 2

Center

Time

Jibe

Activity Selection and Performance Transitiona Average

.29 area mavementb

1.38

.14 1.39

.29 1.64

.17 1.42

2.33

3.93

.41

.34

.19

.27

Social Interaction Average

peer duration0

Frequency of peer

.07

.15

.03

.lO

8.73”

.90

.68

.52

.92

.2 1

1 1.96”

.04

.07

.05

.13

1.36

4.37

1.27

.59

.57

.9 1

1.07

14.54”

2.73

4.32

1.47

.19 19.55”’

contact& Average

adult durationa

Frequency of adult contact+

Co-occurrence of Behaviors Peer-Adult0

.004

.Ol

.oo 1

.02

.44

12.62”

3.46

Peer-Activity0

.09

.08

.o 1

.07

3.34

1.54

2.76

Adult-Activitya

.o 1

.03

.06

.13

9.70”

8.59”

2.43

Peer-Adult-Activity0

.002

.o 1

.oo

.02

.67

16.36”’

3.01

.2 1

1.28

3.23

.5 1

2.68

1.48

3.84

.25

Initiative and Investment lnitiativec Average Activity Duration0 Level of lnvestmentc

1.40

1.55

1.52

1.49

.42

.45

.27

.48

1.56

1.69

1.46

1.54

3.16

aGroup meons represent percent of time observed in these behavioral bAverage occurrence of behavior per two-minute segment. CHigh investment = 2, low investment = 1; initiate = 2, follow = 1. ‘p < .05. l*p < .Ol. l**p < .OOl.

categories.

SOCIAL INTERACTION

209

ported for three of the four co-occurrence behaviors. These main effects were due to children’s spending increased amounts of time in simultaneous involvement with peers, adults, and activities. Significant main effects of time in the predicted direction were also reported for average peer duration. Tendencies (p < .lO) toward a main effect of time in the predicted directions were also found for transition (percentage of time spent in transition decreased), average adult duration (increased), and level of investment (increased). Center Analysis:

Center and Center-by-Time

Effects

As previously described, the frequency, quality, and intent of adult interaction may affect a wide range of children’s behaviors. Table 4 presents an ANOVA of adult behaviors compared by center. An initial matched pairs r-test comparing adult behaviors from time 1 to time 2 revealed no significant changes over time. Therefore, the variables were combined for this ANOVA. Significant center differences were found for the number of adult-initiated contacts, and the percentages of appropriate and restricting behavior. Center B’s adults initiated more contacts, responded more appropriately to child-initiated contacts, and a lower percentage of their childdirected behaviors were restricting compared to Center A’s adults. Further, a tendency (p < .08) was found for children’s initiations toward adults in Center B to be more successful than those initiations of Center A’s children. These differences in adult behaviors directed toward children in the two day care centers, and their supposed impact on children’s behaviors, justified a comparison of centers and center-by-time interactions. As might be expected from the previously reported adult comparisons, Table 3 shows a significant main effect for the number of adult contacts with children (Center B had nearly twice as many contacts as Center A). The differential number of adult contacts seems to be most evident in adults’ involvement with activities of children (i.e., a significantly higher level of children’s simultaneous involvement with adults and activities was found in Center B’s children). TABLE 4 ANOVA on Adult

Behovion

Group Means Variables Adult-initiated

contacts

% Adult-initiated

positive

F

Center A

Center 8

8.00

16.50

28.15”

.87

.95

2.54

4.50

response Child-initiated

contacts

% Child-initiated

positive

4.25

.02

.7 1

1.00

4.88

.92

.98

6.65’

.61

.26

16.72”

response % Appropriate % Restricting

lp l*p

< .05. < .Ol.

response behavior

210

TZELEPIS,

GIBLIN, AND AGRONOW

Duration and frequency of peer contacts also had an interesting interaction when centers were compared. While the average duration of peer contact was significantly greater for children in Center A than those in Center B, the number of peer contacts decreased from time 1 to time 2 in Center A, but increased in Center B during the same period. These results suggested the value of a more detailed and descriptive presentation of patterns of social interactions (i.e., child-child, childadult) existing in the two centers at times 1 and 2. Social Relationship Affilitive Networks. Strayer (1980) describes two levels of analyses by which social interactions may be characterized. First, social relationships may be represented by the frequency of contacts between individuals. In the present study, the two-minute coding procedure provided a record of the specific peers and adults each child contacted every two minutes over the 120 minutes of observation. Following a procedure described by Strayer (1980), the frequencies of the most frequently occurring social relationships (i.e., those peers or adults with whom each child had contact in at least 20% of the total number of two-minute episodes in which they were observed) were compared in a center-by-time ANOVA. The results indicated a strong center-by-time interaction effect (F[l, 141 = 33.80, p 5 .OOl). The mean score decreased from time 1 to time 2 for Center A (time 1 x = 3.13, time 2 x = 2.13), while the opposite trend occurred for Center B (time 1 x = 1.63, time 2 x = 4.00). A second level of analysis of social interactions is one in which the frequencies of social contact are summarized by schematic representations of the groups’ structures (viz., affiliative networks). Figure 1 represents the affiliative networks based on the most frequently occurring interactions for Centers A and B at times 1 and 2. In interpreting these affiliative networks is should be noted that their actual appearance is somewhat arbitrary and determined by a few a priori decisions. Firstly, these networks were composed of both the 16 children who were present in both times 1 and 2, and also other children who enrolled during the first four weeks. Adults who were observed interacting with these 16 target children were also included in these networks. However, a complete record of interactions is presented only for the target children. Secondly, target children were placed in central locations in the affiliative networks. This procedure allowed us to limit the presentation of a target child to a single occurrence in an affiliative network. Finally, since the total amount of time “on camera” varies from one child to another, individuals in a dyad need not spend equal percentages of their available time with one another. For example, in Center A-time 1, child T7 made contact with child T4 40% of all the two-minute segments in which he was observed; T4, however, made contact with T7 in only 20% of his observed two-minute segments. The affiliative networks clearly illustrate the reported significant main effects of center and the significant center-by-time effect. Specifically, Center B’s adults at time 2 had almost twice as many contacts with their children as Center A’s adults had with their children. Interestingly, Center A’s children displayed a decrease from

SOCIAL

211

INTERACTION

h4

04

(4

Figure 1. Affiliative networks for Centers A and Et at times 1 and 2. (a) Center A’s time 1 ofhliative network. (b) Center A’s time 2 afhliative network. (c) Center B’s time 1 offiliotive network. (d) Center B’s time 2 affiliative network. (Subjects are represented by letter and number. ‘7” represents children included in the comparison of times 1 and 2, “A” represents adults, and “P” represents peers of ‘Ts”. Circles represent females and squares represent moles. Lines represent contacts between subjects calculated as the percentage of two-minute episodes in which two subjects were observed interacting: ‘I-” represents 20% contact, “- - - -‘I represents 30% contact, ‘CJ represents 40% contact. Arrows indicate which subiect’s contacts are being described. Arrows pointing away from a subject describe his or her contacts. Absence of arrows indicates on equal percentage of contact between designoted subjects.)

time 1 to time 2 in the number of contacts they had with one another, while an opposite trend occurred in Center B. These findings indicate that the lack of adult contact does not result in an increased contact between children. Rather, high adult contacts with children in their day care center occurs concurrently with an increase in contacts children have among themselves. Still undetermined by these diagrams of affiliative networks is the nature, direction, and extent of adult influence on child behavior. However, this influence may be further described by a sequential analysis of state transitions. Sequential Analysis. Gottman and Bakeman (1979) describe procedures for the sequential analysis of observational data. As they note, recent interest in such procedures arose from research activity on social interaction. This approach seemed particularly appropriate for the present task to assess the influence of the behavior of peers and adults on each child’s behavior. As described, a continuous recording of each child’s involvement with peers,

212

TZELEPIS, GIBLIN,AND AGRONOW

adults, and activities, and the presence or absence of the child from the observed area of the day care center was obtained. Frequency, duration, and overlap of these events have been previously calculated. In the present analysis, these data were transformed into nine mutually exclusive and exhaustive categories which represented an ongoing stream of behavior. These nine categories were contact with a peer or adult (1, 2); involvement with an activity (3); the simultaneous involvement with an activity and peer, an activity and adult, an adult and peer, or an activity, adult, and peer (4-7); off-camera (8); or on-camera but not involved in any of the above behaviors (9). Following procedures described by Gottman and Bakeman (1979), transitions in state from one category to another were recorded. A transition was recorded when a change in prevailing behavior occurred; therefore, a given category of behavior would not follow itself in the sequential record we compiled. Lag-one transitional probabilities were calculated for an entire day care group (Center A or B) at both times 1 and 2. These probabilities were derived from a matrix of lag-one transition frequencies. For example, 20 instances of a child’s concurrent involvement with an activity and an adult were observed in Center B at time 2. Twelve of those events were followed by the child’s continuation of an activity without the presence of an adult. Therefore, the lag-one transitional probability of an activity following a child’s concurrent involvement with an activity and an adult was .60. Figure 2 presents state transition diagrams derived from lag-one transition probabilities. Two criteria determined which behaviors were included in these diagrams: (a) First, categories of behavior were included only if at least 5% of all the transitions observed in that center at that time occurred in that category. For example, in Center B (time 2), only 7 of 155 transitions occurred from involvement with an adult to another state; therefore, this category was not included in the diagram. (b) Second, only lag-one transitional probabilities of .30 or greater are included in these diagrams. Three points summarize the findings illustrated by the state transition diagrams. Firstly, the differential involvement of adults in the activities of children between centers is clearly demonstrated by the number of transitions to and from this state, and the role of this behavior category in the stream of behavior. In Center A the category “activity/adult” occurred so infrequently as to be included only in the diagram of time 2, and at that time only four instances of transition from that were observed. In contrast, in Center B “activity/adult” ocstate to “activity” curred with sufficient frequency to be included in both times diagrams. At both these times “activity/adult” not only predicted the likelihood of a child’s prior involvement in an activity, but also this category was predicted by the child’s prior involvement in an activity. In other words, the adults in Center B fit into the ongoing stream of behavior of their children, while there was no evidence of this occurring in Center A. Secondly, in Center A there was a decrease in the number of transitions from “activity/peer” to “activity”, while in Center B there was an increase in this category. Therefore, the continuing high involvement of adults in children’s activities in Center B seems to encourage the involvement of children

SOCIAL

IOFF CAMERA

I

I

I

ON CAMERA/ UNINVOLVED

IOFF

I

CAMERA/

ON CAMERA/ UNINVOLVED

213

INTERACTION

I

ON CAMERA/ UNINVOLVED

I

1OFF CAMERA

1

I

(ACTIVITY

V4

1

@I

Figure 2. State transition diagrams: Event sequence data. (a) CenterA’s time 1 state transitions. (b) Center A’s time 2 state transitions. (c) Center B’s time 1 state transitions. (d) Center B’s time 2 state transitions. (Numbers in parentheses represent the average number of transitions per day which occurred between designated states for times 1 and 2. “A-B” represents that 40% of all transitions from “A” go to “6”. “A- - -+B” represents that 30% of all transitions from “A” go to “6”~

with one another’s activities. On the other hand, in Center A low levels of adult involvement with children’s activities did not seem to encourage the substitution of peer involvement in place of adult involvement. These results seem in accordance with the decreased complexity of social networks in Center A and the increased complexity of social networks in Center B previously reported (Figure 1). Finally, a high lag-one transition probability between “off-camera” and “activity” in Center B (time 2) indicated the increased tendency of these children to involve themselves

214

TZELEPIS, GIBLIN, AND AGRONOW

more rapidly with activities rather than entering an area is an uninvolved manner. This may indicate either the greater availability of activities in Center B or the continuation of activities begun off-camera. DISCUSSION The present study describes changes in children’s behaviors during the first month of shared day care experience and compares these changes in two day care centers which varied in program orientation. Four categories of behavior measure these changes: activity selection and performance, social interactions, degrees of initiative and investment, and the co-occurrence of activities and social interactions. These four categories of behavior were chosen for their sensitivity to the influence of adult-caretaker behaviors on the adjustments of nascent day care groups to their new settings and their new roles. As predicted, comparisons of the first and fourth weeks of shared day care experience revealed increases in children’s simultaneous involvement with peers, adults, and activities; an increase in the duration of contacts with peers and adults; decreases in the time spent in transition between activities; and an increase in the level of investment in ongoing activities. Interestingly, neither age nor sex significantly affects these behaviors. Comparisons of. the effects of day care centers emphasized the importance of adult behaviors in determining peer social relationships. A dramatic center-bysession effect on the number of peer contacts illustrated the influence of adult contacts. From time 1 to time 2 (weeks 1 and 4 respectively), the number of peer contacts decreased in the center characterized by fewer adult-initiated contacts with children and less appropriate adult responses to children’s overtures. In contrast, the number of peer contacts increased in the same time period in the center with more frequent adult-initiated contacts with children and more appropriate responses to the children’s overtures. Diagrams of affiliative networks further illustrated center differences in peer contacts. Rosenblum and Plimpton (1979) propose that the effects of adults on peer interactions and activities may be direct or indirect. In the present study, diagrams of state transitions seemed to indicate that adults may directly foster the child’s concurrent and subsequent involvement with activities. On the other hand, the frequency and duration of peer contact did not seem to be directly influenced by adults. For example, state transition diagrams did not show a substantial preceding of peer contact by contact with adults. However, one may find support for an indirect effect of adults on peer contacts. In the first week, high peer contact was observed in Center A concurrent with low adult contact. Perhaps in this instance peer contact served as a substitution for low adult contact. However, by the fourth week this same center showed a marked decrease in peer contact. Adult contact remained low. In contrast, Center B, in which adult contact was high in both the first and fourth weeks, showed a significant gain in the frequency of peer social contact and an attendant increase in the complexity of affiliative networks between weeks 1 and 4. Perhaps such concepts as behavioral assimilation or modeling (of the

SOCIAL INTERACTION

215

high levels of adult contact) may be a more appropriate explanation of the long-term effects of adults on peer relationships than substitution (which may be Limited to a short-term influence). The above session and center effects of adult’s contacts on children’s behaviors may be best understood by describing an adult’s role in a nascent preschool group. Generally, the initial orientations of preschoolers are towards adults. This provides adults with opportunities to teach children skills necessary to effectively interact with their peers: to express one’s wants, to understand possession and how to share, to verbalize pleasure and anger, to begin to resolve conflicts. Additionally, adults may redirect children’s attentions to available materials and activities. This may result in instances of parallel play in which adults may encourage children to play interactively. To do so, adults should establish close proximity to children, observe their play so as to anticipate subsequent actions appropriate to interactive play, and concurrently provide insights of each child’s needs to the other. The repetition of such sequences of adult involvement would likely result in an increased co-occurrence of activities with social contacts and an increase in social contacts in the day care setting. These are the results we have reported. In summary, rather than relying upon general labels and impressionistic reporting of adult program orientations (e.g., Montessori, Traditional), we felt it more appropriate to emphasize quantified measures of adult behavior which could more meaningfully be interpreted as affecting the process of children’s responses to an ecological transition. Our findings indicate that children’s activity selections and performance, social interaction and interests, and involvements in their environments are enhanced by an approach in which the teachers are warm, supportive, noncritical, and as nondemanding as possible. Additionally, frequent adult-initiated contacts with children did not decrease peer relationships or hinder independent activity; rather, such adult behaviors fostered these behaviors. REFERENCES Anderson, C. W., Nagle, R. I., Roberts, W. A., & Smith, .I. W. Attachment to substitute caregivers as a function of center quality and caregiver involvement. Child Developmenr, 1981, 52, 53-61. Belier, E. K. Teaching styles and their effects on problem-solving behavior in Head Start programs. In E. Grotberg (Ed.), Critical issues in research to disadvantaged children. Princeton: Educational Testing Service, 1969. Berk, L. E. Effects of variations in the nursery school setting on environmental constraints and children’s modes of adaptation. Child Development. 1971, 42. 839-869. Billman, J., & McDevitt, S. C. Convergence of parent and observer ratings of temperament with observations of peer interactions in nursery school. Child Development, 1980, 51, 395-400. Bronfenbrenner, U. The ecology of human development: Experiments by nature and design. Cambridge, Mass.: Harvard University Press, 1979. Fagot, B. Influence of teacher behavior in the preschool. Developmental Ps.ychology. 1973, 9, 198-206. Feldbaum, C. L., Christenson, T. E., & O’Neal, E. C. An observational study of the assimilation of the newcomer to the preschool. Child Development, 1980, 51. 497-501. Gottman, J. M., & Bakeman, R. The sequential analysis of observational data. In M. E. Lamb, S. J. Suomi, & G. R. Stephenson (Eds.), Social interaction analysis: Methodological issues. Madison, Wis.: The University of Wisconsin Press, 1979.

216

TZELEPIS,

GIBLIN, AND AGRONOW

Huston-Stein, A., Friednch-Cofer, L., & Susman, E. J. The relation of classroom structure to social behavior, imaginative play, and self-regulation of economically disadvantaged children. Child Developmenr. 1977, 48, 908-916. McGrew, W. C. Aspects of social development in nursery school children, with emphasis on introduction to the group. In N. B. Jones (Ed.), Ethological studies of child behavior. Cambridge: Cambridge University Press, 1972: Miller, L. B., & Dyer, .I. L. Four preschool programs: Their dimensions and effects. Monographs of the Society for Research

in Child Development,

1975, No. 162.

Rosenblum, L. A., & Plimpton, E. H. The effects of adults on peer interactions. In M. Lewis & L. A. Rosenblaum (Eds.), The child and its family. New York: Plenum, 1979. Stodolsky, S. S. How children find something to do in preschools. Genetic Psychology Monographs, 1974,

Strayer,

90, 245-303.

F. F. Social ecology cognition,

of the preschool

peer group.

affect, and social relations (The Minnesota

In W. A. Collins (Ed.), Development of Symposia on Child Psychology, vol. 13).

Hillsdale, N.J.: Lawrence Erlbaum, 1980. Thompson, Cl. G. The social and emotional development of preschool children under two types of educational programs. Psychological Monographs. 1944, 56. (5, whole no. 258). Winer, B. J. Sratistical principles in experimental design. New York: McGraw-Hill, 1971.