Potential impact of classroom computer use on preschool children’s social behaviors


Dialog, 17(3), 16-34 

Copyright © 2014,  

ISSN: 1930-9325 

 

RESEARCH ARTICLE 
 

 

 

Impact of Classroom Computer Availability on Preschoolers’ Social 

Interactions 
 

 

Heather C. Janisse 

Eastern Michigan University 

 

Navaz Bhavnagri, Xiaoming Li and Bonita Stanton 

Wayne State University 

 

Sara Johns 

Eastern Michigan University 

 

Sheretta Butler-Barnes 

Washington University in St. Louis 

 

 
Despite increases in computer availability in the classroom, very little data exists on 

computer availability and its use in Head Start classrooms. Therefore, the purpose of this 

study was to examine the impact of availability of computers on urban preschoolers’ 

naturally occurring social interactions in Head Start classrooms across one school year. 

The sample consisted of 66, predominantly African American, children in four urban, 

Head Start classrooms. A quasi-experimental design was employed in which three 

classrooms had computers, while one classroom served as the control. Naturally 

occurring social interactions were observed during free-play sessions over an 8-month 

period. Hierarchical linear modeling was used to analyze differences in social interactions 

across classrooms and school year. Results suggest that classroom computer availability 

may increase the interactive behavior of preschoolers. Results suggest that investing 

resources into a computer center in the Head Start classroom can have a positive effect on 

social development. 

 

 

Keywords: social interaction, computers, African American  

 

 

Advancement of educational technology over the past two decades has resulted in government 

policy to increase computers in the classroom (McMillan Culp, Honey, & Mandinach, 2003). 

Children now get an early start on becoming technologically literate (e.g., No Child Left Behind 

Act, 2001; NAEYC, 2012). Therefore, computers are increasingly being integrated into young 

children’s classrooms.  



COMPUTERS AND SOCIAL INTERACTION     17 

 

In 1987, only 15% of the surveyed preschool programs in one Midwestern county had 

computers available for children (Donohue, Borgh, & Dickson, 1987). However a decade later, 

80-90% of early childhood educators reported availability of computers in the classroom 

(Haugland, 1997; National Center for Educational Statistics [NCES], 2003). A random sample of 

Head Start teachers in one state found that 88% had at least one computer available inside the 

classroom to use with children (Bewick, 2000). The ratio of computers to students has changed 

from 1:125 in 1984 to 1:22 in 1990 (Clements & Swaminathan, 1995) to 1:6 in 1999 (NCES, 

2003), thus indicating that their availability for use has increased rapidly.  

 Despite this increase, there is a substantial disparity in access to and functionality of 

technology between different socio economic status (SES) niches. Judge, Pucket, and Bell (2006) 

found that 96.5% of upper-income households had a computer for child use, while only 45.5% of 

lower-income households had a computer. Regardless of SES, Caucasian parents were more 

likely than African American or Latino parents to have Internet access (Calvert, Rideout, 

Woodlard, Barr, & Strouse, 2005). Furthermore, kindergarteners from minority or low SES 

families are less likely to attend schools that provide students with access to the Internet (Parsad 

& Jones, 2005; Rathbun, West, & Hausken, 2003) and the ratio of students to computers was 

higher in schools with higher poverty concentrations (5.1 to 1 versus 4.2 to 1, respectively; 

Parsad & Jones, 2005).  

 Despite the general increases in computer availability in the classroom, very little data 

exists on computer availability and its use in Head Start classrooms (Bewick, 2000). According 

to Bewick (2000), the majority of Head Start teachers in one state reported that they had 

computers available for use with children. However, 17% said they didn’t use them at all with 

the students. The majority (68%) had only one computer for a classroom of 16-20. Furthermore, 

lack of incorporation of the computers into the curriculum, inadequate equipment, and some 

concern over the appropriateness of use with young children was noted by some respondents. 

Therefore, the purpose of this study is to examine the impact of availability of computers on 

urban preschoolers’ naturally occurring social interactions in the Head Start classroom across 

one school year. 

 

 

MAJOR ISSUES ON COMPUTERS AND SOCIAL INTERACTIONS  
 

The following review examines the issues related to computers and social interactions that have 

arisen over the past few decades. We aim to provide a new direction and shift the old debates on 

this topic. 

 

 

Are Computers Developmentally Appropriate Materials? 
 

The first issue is the debate among educators and researchers regarding the value of personal 

computers in early childhood development (Clements & Samara, 2003; Cordes & Miller, 2000; 

Haugland & Wright, 1997; Lepper & Gurtner 1989; National Association for the Education of 

Young Children [NAEYC], 1996; Scoter, Ellis, & Railsback, 2001). The major debate in the 

1990’s was whether computers were developmentally appropriate for young children (Clements 

& Nastasi, 1993; NAEYC, 1996; Shade & Watson, 1990) and if so, could young children operate 

the computers (Borgh & Dickson, 1986; King & Alloway, 1992).  



18    JANISSE ET AL.  

 
 

To resolve this debate, computer proponents have provided empirical evidence that 

computers, when used appropriately and interactively, have the potential to transform 

conventional material for children, making it easier to organize and access information (Freeman 

& Somerindyke, 2001) and can benefit children’s cognitive and social development (NAEYC, 

2012). Moreover, they have found that young children can use computers; activate them, follow 

pictorial directions, and use cues to reason about their activity (Clements & Nastasi, 1993; Hess 

& McGarvey, 1987). However, some educators have cautioned that assumptions about children’s 

cognitive development cannot be made based on children’s proficiency in manipulating 

computer icons (e.g., Elkind, 1996). 

 

 

Do Computers Cause Social Isolation in children? 
 

The second issue in the 1980’s was a concern that computers in the classroom would cause 

children to be isolated from their peers and deprive them of socialization (Baker, 1985; Barnes & 

Hill, 1983; Lipinski, Nida, Shade, & Watson, 1986). This notion of “social interaction 

deprivation” by computer has not been supported in the research literature. Typically, children 

use computers in dyads or groups in classroom settings (e.g., Clements, 1994; Swigger & 

Swigger, 1984). Hence a plethora of social interactions are reported.  

For example, Heft and Swaminathan’s (2002) reported preschoolers observing and 

acknowledging each other; commenting, sharing and helping each other on the computer; as well 

as peer conflicts regarding turn-taking and sharing of the computers. Bergin, Ford and Hess 

(1993) reported that 95 kindergartners cooperated and took turns while on the computer. They 

observed only two aggressive behaviors during their four month study. Shahramin and 

Butterworth (2002) reported the most frequently coded interactions were directing partner’s 

actions, providing and asking for information, self-monitoring, declarative planning, showing 

pleasure, and disagreeing with partner on the computer.  

 

 

How Does Computer Activity Compare to Other Activity? 
 

A third issue is to compare the social interactions when working on a computer, which is an 

object, to social interactions when playing with other objects. Some researchers suggest that 

young children show increased pro-social behaviors when working together on a computer 

compared to other classroom activities such as toy-focused play (Lipinski et al., 1986; 

McCormick, 1987; Svensson, 2000). Muller and Perlmutter (1985) found that preschoolers spent 

significantly less time engaging in peer interactions while at puzzle play (7%) than they did 

while at computer play (63%). Additionally, only 11% of the time was spent in solitary activity 

at the computer, while 55% of the time was spent in solitary activity with the puzzle. Anderson 

(2000) found that 4-year-old’s cooperative play in the computer center paralleled the proportion 

of cooperative play in the block center. It also provided a context for sustaining interaction that 

could be transferred to play in other areas as well. These studies cumulatively suggest that 

availability of computers as objects solicits social interactions in children. 

 

 

 

http://firstsearch.oclc.org/WebZ/FSQUERY?searchtype=hotauthors:format=BI:numrecs=10:dbname=PsycINFO_1887::termh1=Svensson%5C%2C+Ann-Katrin:indexh1=au%3D:sessionid=fsapp8-36408-eu1fp4fw-r95ice:entitypagenum=4:0:next=html/records.html:bad=error/badsearch.html


COMPUTERS AND SOCIAL INTERACTION     19 

 

Do Computers Displace Important Activities? 
 

A fourth issue is whether computer activity displaces other important developmental activities. 

Hohmann (1994) argues that for preschoolers and kindergarteners, the addition of computers to 

the environment has positive social consequences and appears not to disrupt other classroom 

social interactions. Lipinski et al. (1986) found that computer novelty initially interrupted free-

play activity patterns by drawing children away from traditional activities. However, most 

activities returned to baseline levels within two weeks. Other research has supported this finding 

(e.g., Bergin et al., 1993; Clements & Nastasi, 1985). These studies suggest the hypothesis that 

the presence and use of computers does not adversely impact the behaviors occurring during 

other classroom activities. 

 

 

LIMITATIONS OF PAST RESEARCH  
 

The studies reported above have addressed many important questions regarding computer use in 

early education. However, these studies have several limitations. First, all studies examined the 

social interactions of children when they are on the computer and not the impact of availability 

of computers on social interactions during free play in the classroom. Second, no known studies 

have used a design in which they compare the social interactions of classrooms with and without 

computers. Third, most of these studies had predominantly middle class and European-American 

children (e.g., Baker, 2000), and not low-income, inner city, and/or minority children who have 

substantially less access to computers (Calvert et al., 2005; NCES, 2003). Fourth, most of the 

studies presently available involve small sample sizes, usually of only one or two preschool 

classrooms (e.g., Freeman & Somerindyke, 2001). Fifth, some studies did not operationally 

define specific behaviors such as sharing (Muller & Perlmutter, 1985) and cooperative play 

(Anderson, 1998). Sixth, most studies have examined computers in a pre-prescribed social 

context such as working together, working separately, or performing certain predetermined tasks 

(e.g., Shahramin & Butterworth, 2002). 

Our study addresses these limitations by (1) examining the influence of availability of the 

computer on social behavior during free play; (2) having a contrast group; (3) examining lower 

income children in a community setting; (4) having a more adequate sample of four classrooms; 

(5) having operational definitions of all observed behaviors; and (6) examining social 

interactions in a naturally occurring classroom environment. 

 

 

GOALS OF CURRENT STUDY 
 

Given the above issues and limitations, our goal was to open up a new line of investigation; 

namely, to examine the impact of availability of computers on the social interactions of urban, 

low-income preschool children during indoor free play in classrooms. 

Therefore our research questions are as follows: How does the availability of computers 

in classrooms impact 3 to 5- year-olds’ (1) overall social interactions during indoor free play, (2) 

developmental trajectories of social interactions during indoor free play across the school year 

and (3) social interactions expressed verbally, the affect displayed during these interactions as 



20    JANISSE ET AL.  

 
 

well as the target of these social interactions? These questions are largely exploratory because 

the vast majority of past research has examined child activity while on the computer only.  

However, based on the research reviewed above and other studies that have found 

positive effects of computers on child development (e.g., Rhee, & Bhavnagri, 1991; Li & Atkins, 

2004; Li, Atkins, & Stanton, 2006; Floyd, Canter, Jeffs, & Judge, 2008; Lonigan, Allan & Lerner, 

2011; Diamond & Lee, 2011), we made the following conservative hypothesis: The availability 

of computers in the classrooms would not have a significant adverse impact on preschoolers’ 

social interactions during indoor free play. 

 

 

METHOD 
 

Participants 
 

The study sample consisted of 66 preschoolers enrolled in a Head Start Program in a Midwest 

urban city. Participants were drawn from a larger study examining the effects of computers on 

the development of Head Start children. Children were recruited from four classrooms at three 

Head Start sites. The mean age of the participants at the beginning of the school year was 48.71 

months (S.D. = 6.59). The majority of the children were African American (95.5%), while 1.5% 

of the children were Caucasian, and 3% were identified as other. Annual household income 

ranged from below $5,000 to more than $50,000, with 50.9% of parents reporting their income 

was between $0-10,000 per year. The majority of mothers (78.6%) and fathers (77.5%) had at 

least a high school diploma or GED; 11.4% of mothers and 6.9% of fathers had a college degree. 

See Table 1 for details of socio-demographic characteristics by classroom. Classrooms did not 

differ significantly on any variables detailed in Table 1. All parents completed a written 

informed consent and agreed to let their child participate in the study. 

 

 

TABLE 1 
Sample Characteristics 

 Computer 

Room 1 

Computer 

Room 2 

Computer 

Room 3 

Control 

Room 4 

N 17 16 18 15 

Boys 8(47%) 7(44%) 13(72%) 9(60%) 

Age in months 48.76(6.41) 48.06(6.72) 47.11(6.99) 50.13(6.30) 

African American  94.1% 87.5% 100% 100% 

Household income ≤$10,000 46.7% 50.0% 71.4% 57.1% 

Mother <HS education 35.3% 20.0% 23.5% 33.3% 

Father <HS education 35.3% 21.4% 21.4% 41.7% 

Mother work full-time 58.8% 37.5% 70.6% 33.3% 

Father work full-time 37.5% 53.3% 63.6% 50.0% 

Access to home computer 24% 62% 44% 40% 

Total # of observations  1046 1045 1130 1079 

Mean child observations/day  2.78(.40) 2.89(.10) 3.22(.10) 3.12(.15) 

Mean observations/child 61.53(32.67) 71.93(31.39) 65.31(18.54) 62.78(21.68) 

 



COMPUTERS AND SOCIAL INTERACTION     21 

 

Study Design  
 

The study used a quasi-experimental design in which self-contained classrooms were randomly 

assigned to either a computer or control condition. Only classrooms that were considered typical 

and had not previously used computers were considered for inclusion.  

The control and computer classrooms were similar in child demographics, teacher 

characteristics, and classroom climate. As stated above, children did not differ significantly on 

child/family characteristics. All teachers were African American, had a Child Development 

Associates degree, and followed standard Head Start curriculum. Informal observations suggest 

that teachers also had similar styles with regards to classroom guidance and discipline.  

With regard to general classroom climate, there were many similarities because all 

classrooms were members of the same Head Start Agency. All classrooms met for a half-day 

session and had one teacher and one teacher’s assistant present during class time. The physical 

arrangement and use of space was similar, including the availability of the same learning centers 

in each classroom. Classrooms followed the same curriculum and daily routine. Both the control 

and computer classrooms allowed for one hour of free-play in classroom learning centers and 

therefore provided opportunities for child interaction. Lastly, classrooms shared the same 

educational philosophy, based on developmentally appropriate practices by the NAEYC and 

followed the same rules governing classroom conduct and usage of equipment. 

Four classrooms were selected from the larger study to participate in this observational 

study. Three of these classrooms had computers. The fourth classroom had no computer and 

served as the control. Treatment classrooms had two computers in each room. The “twin 

imagination station” (TIS) developed by Hatch, Inc (Winston Salem, NC) was installed in each 

of the computer classrooms. The TIS is a child friendly hardware system that includes two 

computers. The computers were pre-installed with age-appropriate educational software 

programs (e.g., Millie's Math House and Bailey's Book House) and were set up near one another 

with chairs for two children per computer.  

 

 

Procedure 
 

As part of the larger study’s protocol, each child in a computer classroom had approximately 15 

minutes per day, as part of their daily curriculum, to play on the computer with their choice of 

developmentally appropriate educational programs. Head Start teachers and/or graduate research 

assistants provided necessary assistance during the first couple of weeks to familiarize children 

with the login process and the operation of the mouse, keyboard, and printer. However, there was 

minimal adult involvement beyond the initial period. Although literature suggests that teacher 

involvement can greatly enhance student learning on the computer (Judge et al, 2006; Plowman 

& Steven, 2005), typical resources available in the Head Start classroom does not allow for one-

on-one time with each child every day on a computer. Therefore the current study was designed 

to be ecologically valid with the goal of examining the impact of computer center availability in 

the classroom. Children’s computer use was based on either self-selection of computer for play 

or direct solicitation of the teacher to take a turn on the computer if desired. Potential computer 

use occurred throughout the school day and was available for children to use during their free-

play hour. Children were allowed to bring another child to the computer station to work together 

if desired. No specific teacher instruction was given to students regarding computer use. Other 



22    JANISSE ET AL.  

 
 

activities available during the free-play hour included the use of standard Head Start learning 

centers such as literacy, art, manipulative, dramatic play, and block centers.  

The control classroom received the standard Head Start curriculum including the 

availability of the standard learning centers listed above. Live observations of social interactions 

were conducted in the Head Start classrooms during each classroom’s one hour free-play session. 

The free-play session was chosen for observations because it typically provides optimum 

opportunities for self-selected social interactions. During free-play, the child could have been in 

the computer center or any other center in the classroom. Therefore, observations occurred 

across varied contexts. Observation sessions occurred twice a week, every other week for eight 

months (October through May), allowing for approximately 60 observations per child. Table 1 

shows average number of observations overall and per day by classroom. Sixty observations 

were considered adequate based on previous studies using this methodology for observing peer 

interactions in classrooms (Bhavnagri, 1987; Ladd, 1981). Variation in number of observations 

per child (as shown in Table 1) occurred due to differences in attendance, drop-out and 

availability of each child. Classrooms did not differ significantly on number of observations per 

child. 

 

 

Coding 
 

Each child was observed using an event sampling method. The specific procedure used for this 

event sampling is called scan sampling. This approach is borrowed from the field of ethology 

where the social behaviors of a group of animals are observed in a natural setting. Each child was 

observed for about 10 seconds, which was adequate to determine the specific details regarding 

social interactions and then the observations were recorded on a coding sheet immediately. Next, 

the coder would scan the room looking for the next child to be observed. 

The observers used a structured coding system developed by Bhavnagri (1987). This 

coding system was designed specifically for examining young children’s social competencies in 

early childhood educational settings and is representative of naturally occurring classroom 

interactions. All children’s social interactions, (i.e., inclusive of social play behaviors and other 

social behaviors which were not play) were recorded. These behavioral interactions were 

expressed simultaneously through three channels of communication, namely through their 

physical social interactions, labeled Social Interaction and verbal interaction labeled Verbal 

Interaction and through their affective communications labeled Affect. The Target to whom these 

behaviors were directed was also specified. Refer to Table 2 for operational definitions of the 

observational coding categories. The Social Interaction code, which must include some physical 

behavior, is considered the primary behavioral interaction category. The Verbal, Affect, and 

Target codes are considered supplementary categories and are treated as categorical. This coding 

scheme has demonstrated good reliability for use with preschool children. In an earlier study, the 

reliability ranged from .71 to 1.00 (Bhavnagri, 1987). 



COMPUTERS AND SOCIAL INTERACTION     23 

 

TABLE 2 
Coding Scheme 

Behavioral Categories Description 

I. Social Interaction 

 

 

1. Isolated Behavior No verbal or physical interaction with anyone, but child 

interacts physically with objects and self. No interest in peer is 

shown. 

2. Passively Observes Watches child, teacher, or peer group without any other social 

behavior 

3. Parallel Play Plays side by side with child, teacher or group with similar objects 

or actions, engaged in similar activity 

4. Positive Interactions Shares, takes turns, simultaneously acts upon toys, initiates, 

cooperates, joint activity with child, teacher or group 

5. Negative Interactions Hitting, punching, shoving, grabbing, pulling hair of child, teacher 

or group 

II. Verbal Interaction 

 

 

1. Positive Verbal Statements Verbal content is pro-social, e.g., requests and praises, agrees on 

goals, offers turns, encourages, suggests cooperation. 

2. Negative Verbal Statements Verbal content is antisocial, e.g., expresses aggressiveness, 

rejection, annoyance, refusal to cooperate, whines, complains 

3. Neutral Verbal Statements Verbal content is not related to social skills, e.g., talks about 

properties of objects, talks or reads to self 

4. Not Applicable Child is not engaged in any verbal activity 

III. Affect 

 

 

1. Positive Affect Smiles, laughs, giggles, claps, bounces, hugs or kisses 

2. Negative Affect Cries, sulks, frowns, fusses, acts agitated, pouts 

3. Neutral Affect No affect displayed 

IV. Target of Interaction  

 

 

1. Adult  Teachers, parents, volunteers, observers 

2. Boy Boy 

3. Girl Girl 

4. Peers A group of children 

5. Self Alone with no object 

6. Object Toy or object 

 

 

Inter-coder agreement.     An expert in the observational scan sampling method trained 

the coders in using this coding system. The expert also monitored the coders’ observations when 

they started collecting data at the Head Start centers to address any issues. Twenty percent of 

total observations were coded independently by two trained graduate research assistants in order 

to establish inter-rater reliability. Cohen’s kappa (Cohen, 1960)
 
was used as a measure of inter- 

coder reliability. For this sample, kappa was equal to .95 for Social Interaction, .87 for Verbal 

Interaction, .93 for Affect, and .91 for Target. 



24    JANISSE ET AL.  

 
 

Statistical analysis  
 

Hierarchical linear modeling (HLM; Bryk & Raudenbush, 1992; Raudenbush & Bryk, 2002) was 

used to address the first two research questions of this study (How does the availability of 

computers in classrooms impact 3 to 5- year-olds’ (1) overall social interactions during indoor 

free play, (2) developmental trajectories of social interactions during indoor free play across the 

school year.) HLM methodology is ideally suited for the current study given the nested nature of 

the data (longitudinal data collected from children nested within classrooms). Taking the nested 

nature of the data into account is important for several reasons. Children within a particular 

classroom share the same teacher and physical environment and have similar learning 

experiences, which may lead to increased homogeneity over time (Osborne, 2000). Therefore, 

observations are not fully independent, which violates a primary assumption of most analytic 

techniques and increases Type I error (Tabachnick & Fidell, 2007). Hence, HLM was 

implemented through HLM 6.0 (Raudenbush, Bryk, Cheong, & Congdon, 2004). The dependent 

variable used in this analysis was Social Interaction. 

In order to answer research question 3 (Are there differences in verbal interaction, affect, 

or the target of social interactions based on computer availability?) analysis of covariance 

(ANCOVA) was employed. The broad categories of Verbal Interaction, Affect, and Target were 

transformed into variables that represent the proportions of all observed interactions across the 

school-year that fell into each possible sub-category reflected in the coding scheme described in 

Table 2 (e.g., proportion of interactions that included positive affect, proportion with negative 

affect, and proportion with neutral affect). One proportion variable for each sub-category was 

obtained for each child. Therefore these analyses did not include a repeated measures component 

and had a total sample size of 66 children. HLM was not utilized for these analyses because the 

sample size was insufficient (66 children nested within four classrooms).  

Variables were screened for univariate and multivariate outliers. Five univariate outliers 

were found on proportion scores and winsorized. No extreme skewness was detected; therefore 

all variables were left untransformed to increase interpretability of findings. 

 

 

RESULTS 
 

Preliminary Analyses 
 

Preliminary analyses revealed that the three computer classrooms were not significantly different 

from one another on Social Interaction, Verbal Interaction, Affect, or Target. Therefore, 

classroom was assigned to control (0) versus computer (1) for further analyses. There continues 

to be no significant differences on measured demographic or classroom characteristics from 

Table 1 based on this reassignment of classroom.  

For HLM analysis, Social Interaction was treated as an interval level outcome. Isolate 

play was given a value of 1, passively observing was given a value of 2, and parallel play was 

given a value of 3. Positive and negative social interactions were combined because they both 

reflect that there was the presence of a socially interactive behavior, be it positive or negative, 

and given a value of 4. It should be noted that negative social interactions were relatively few 

(4% of total behavior). Therefore the Social Interaction outcome used in the HLM analysis 

reflects a continuum of social interactions on a scale of 1 to 4. Past literature has supported the 



COMPUTERS AND SOCIAL INTERACTION     25 

 

conceptualization of these behaviors as correlated to maturity and therefore ordered (e.g., Rubin 

& Krasnor, 1980; Smith, 1973).  

To account for changes across time, a time variable was created ranging from -240 days 

to zero. Time was calculated this way so that the last observation of the year was zero. Hence, 

significance tests of effects of the intercept would reflect differences at the end of the school year, 

taking start values into account. This allowed for testing general differences between classrooms 

on Social Interaction while including the time effect in the model. The average number of 

observations per student was 65.15 (Range = 4-107, S.D. = 26.32, Median = 71.0). 

 

 

Hierarchical Model of Social Interaction 
 

The current data structure represents a three-level model. First level units were individual 

observations of social interactions, resulting in a total of 4300 observations for analysis. Second-

level units were the 66 children enrolled. The four classrooms comprised the third-level units. 

The small number of classrooms would not allow for analysis of the classroom effect at level 

three. Therefore, the classroom effect was entered as a covariate in the level-two model. 

 A full model was analyzed that allowed for an estimation of the effects of child and 

classroom predictors on the shape of the trajectory of Social Interactions (i.e., physically 

communicated social interactions). The focus was on assessing the importance of child 

characteristics including gender, age, and family income and classroom assignment. The model 

is represented as follows: 

 

Status at end of school year: 

 

π0 = β 00 + β 01*(INCOME) + β 02*(GENDER) + β 03*(AGE) + β 04*(CLASS) + r0 

 

Growth Rate/Slope: 

 

 π1 = β 10 + β 11*(INCOME) + β 12*(GENDER) + β 13*(AGE) + β 14*(CLASS) + r1 

 

The constants (β00 and β10) in the equations define the status of the outcome variable at 

the end of the school year and the growth curve of the outcome when all other variables in the 

equation equal 0. Gender and class assignment were left un-centered. Income and age were 

centered on the grand mean. Therefore, the coefficients for gender and classroom assignment 

represent differences between groups at the end of the school year and differences between the 

groups on growth rate. The coefficients for age and income represent changes in status at the end 

of the school year and on growth rate as a function of income and age. Because the equation for 

the growth rate reflects change across time, the variable coefficients in that equation represent 

the interaction between model variables and time (e.g., class assignment by time). 

The full model as a whole was significantly better than the null model in which only the 

intercepts were included, χ
2

diff(11, N=4300) = 14611.95-14538.97 = 72.98, p < .001. Thus the 

predictors as a group improved the model beyond that produced by chance. Class assignment 

was a significant predictor of status at the end of the school year and there was a significant 

change in outcome across time for all groups. All other predictors were non-significant.  



26    JANISSE ET AL.  

 
 

Therefore, a reduced model was proposed in which only predictors with p values < .20 

were included in the model. In this model, class assignment and gender were included as 

predictors of end of school year status (age and income were omitted) and gender and age were 

entered as predictors of growth rate (class assignment and income were omitted). This reduced 

model did not differ significantly from the full model, χ
2
diff (4, N=4300) = 14539.72-14538.97 

= .75, p > .05. Estimated fixed and random effects for this model are presented in Table 3.  

Results show that classrooms (control versus computer) differed on end of school year status. On 

average, computer classrooms showed significantly higher levels of social interaction at the end 

of the school year. Computer classrooms had a .44 point higher level of social interaction than 

the control classroom on a scale of 1 to 4. Post hoc analyses revealed that this difference is 

largely due to a higher proportion of positive social interactions in the computer classroom (34% 

for computer room versus 19% in control room). 

Table 3 also shows that there was a significant effect of time (intercept for growth rate) 

across all groups for Social Interaction. Both classroom conditions showed a significant increase 

in level of Social Interaction across the school year. This suggests that, in general, children were 

likely to show increasingly interactive behavior and play as the school year progressed.  

No statistically significant associations were found for gender, age, family income and 

Social Interaction. That is, there is no evidence that level of Social Interaction was generally 

related to the gender, age, or family income of the child or that the trajectory of Social 

Interaction across the school year was influenced by these child demographics. 

 

 

TABLE 3 
Final HLM Results for Social Interaction 

Fixed Effect Coefficient SE t-ratio p value 

Model for end of school-year status     

 Intercept 2.22 0.11 19.96 <.001 

 Gender
a
  -0.21 0.11 -1.83 .071 

 Class
b
  0.44 0.11 4.20 <.001 

Model for growth rate 
    

 Intercept 0.01 0.00 2.68 .010 

 Gender  -0.001 0.00 -1.58 .119 

 Age  -0.0001 0.00 -1.64 .105 

Random Variance Coefficient  df p value 

 Level 1 variation 1.68    

 Level 2     

  End of school-year status (r0) 0.08  57 <.001 

  Growth rate (r1) 0.00  60 <.001 

 Level 3 Intercept 0.002  3 .136 
a
 0=male, 1=female; 

b
 0=control, 1=computer; Effects non-significant at p ≥ .20 were not retained in the final model. 

 

 

 

 

 



COMPUTERS AND SOCIAL INTERACTION     27 

 

ANCOVA Analyses 
 

ANCOVA was employed to examine differences in proportions of behaviors observed for Verbal 

Interaction, Affect, and Target of social interactions separately. Observed proportions for each 

behavioral category by classroom can be seen in Table 4. Because of the small sample size for 

these analyses (N=66), only demographic characteristics and two-way interactions that were 

significant at p < .20 were included in the final models and presented here. Table 4 also presents 

the Univariate F-test results for these ANCOVA analyses.  

 

 

TABLE 4 
Proportion Behaviors Observed (Standard Deviation) by Classroom Condition and 

Univariate F Statistics from ANCOVA Analyses (N=66) 

 Proportion Observed 

(S.D.)                                                                      

Effect 

 

  

Control 

 

Computer 

 

Class
a 

 

Gender
b 

 

Age 

 

Income 

Class x 

Gender 

 

Verbal Interaction 

 

       

1. Positive .08 (.08) .09 (.08) .26 n/a n/a 2.77 n/a 

2. Negative .01 (.02) .03 (.04) 5.79* n/a n/a n/a n/a 

3. Neutral .18 (.08) .26 (.09) 11.86** 2.34 2.36 n/a n/a 

4. Not Applicable .73 (.13) .62 (.14) 9.28** n/a n/a 3.15
†
 n/a 

Affect 

 

       

1. Positive .18 (.13) .21 (.09) 2.75
†
 5.83* n/a n/a 8.23** 

2. Negative .02 (.02) .04 (.04) 6.74* n/a n/a n/a n/a 

3. Neutral .80 (.12) .74 (.10) 7.45** 5.48* n/a n/a 8.47** 

Target 

 

       

1. Adults .14 (.07) .12 (.05) 2.51 9.31** n/a n/a 2.77 

2. Boy .12 (.07) .15 (.08) 4.82* 28.95** n/a n/a n/a 

3. Girl .07 (.04) .12 (.09) 4.95* 23.24** n/a 3.68
†
 n/a 

4. Group .11 (.05) .13 (.07) .88 8.54** n/a 3.33
†
 n/a 

5. Self .03 (.02) .01 (.01) 7.79** n/a n/a n/a n/a 

6. Object 

 

.54 (.11) .46 (.14) 3.76
†
 n/a n/a n/a n/a 

† 
p < .10, * p < .05, ** p < .01; 

a
 0=control, 1=computer; 

b
 0=male, 1=female; n/a = not entered based on p < .20 

criterion 

 

 

For verbal interaction, computer classrooms displayed significantly more verbal 

statements that were neutral in content (e.g., verbal content that focused on topics such as 

properties of objects, talking or reading to self. See operational definitions, Table 2 for details), 

some negative statements and significantly less non-verbal interactions (see Table 4). This 



28    JANISSE ET AL.  

 
 

indicates that children in the computer classroom condition talked significantly more than the 

control group, as indicated by significantly less non-verbal interactions. Furthermore, it informs 

us about the verbal content of their conversations when compared to the control group. They 

talked significantly more about non-social topics (i.e., neutral statements). They also had more 

negative communications, such as expressing aggressiveness, annoyance, or complaints, 

although these negative communications were relatively few (3%). 

Results for the affect showed that the computer classrooms displayed significantly more 

negative and less neutral affect. There was a trend toward more positive affect displays by 

children in the computer classrooms (p < .10). There was a main effect of gender for positive and 

neutral affect. Girls displayed significantly less positive affect and more neutral affect. The class 

by gender interaction was also significant for positive affect and neutral affect. Post hoc analyses  

show that boys in the control classroom displayed more positive affect and less neutral affect 

than girls. Also, girls seem to carry the effect of classroom assignment. It was girls in the 

computer classroom condition who showed significantly less neutral affect, more negative affect, 

and a trend toward more positive affect.  

 Results for the target of behavior proportions show that computer classrooms were 

significantly more likely to target behaviors toward boys and girls, and less likely to target self. 

This finding suggests the children in the computer classroom targeted interactions with persons, 

rather than playing alone. Girls were more likely to target behavior toward adults and other girls 

and less likely than boys to target other boys or groups of peers. This suggests that both girls and 

boys preferred to interact with other peers of their own sex. There was no significant difference 

on targeting adults or groups based on computer availability. 

 

 

DISCUSSION 
 

Overall, our hypothesis that the presence of computers in the classroom would not have an 

adverse impact on preschoolers’ social interactions was largely supported. Children in 

classrooms with computers did not show lower amounts of interactive behavior across social, 

verbal or affect domains. On the contrary, the children in computer classrooms were more likely 

than children in classrooms without a computer to display socially interactive behaviors, to be 

verbal, and to display more affect. Children in the computer classroom condition were also more 

likely to interact with peers and less likely to play alone.  

There was some evidence that negative social, verbal, and affective interactions were 

more prevalent in the computer classroom. However, negative interactions were infrequent (≤ 

4%) and this difference may be a consequence of increased social interactions. Both positive 

interactions and negative ones are likely to increase across the school-year (Ramsey, 1995) and 

friends are likely to engage in more cooperation and conflict than non-friends (Hartup & Brett, 

1987; Hartup, 1992). Because children are interacting with one another more frequently, it 

increases the chances of positive and negative behaviors. We do not interpret this as an 

indication that computer presence is harmful in light of the other findings.  

Both computer and control classrooms displayed an increase in social interactions across 

the school year. This finding is fitting with developmental expectations for children of preschool 

age (Rubin & Krasnor, 1980; Smith, 1978) and therefore provides additional support for the 

validity of the coding scheme used in this study. 



COMPUTERS AND SOCIAL INTERACTION     29 

 

Together, the results from this study suggest that the availability of computers in urban 

preschool classrooms may have a positive influence on children’s social interactions. This 

conclusion is consistent with past studies that have found the interactions of children on the 

computer to be interactive and social in nature (e.g., Bergin et al., 1993), perhaps more social 

that other classroom activities (e.g., Muller & Perlmutter, 1985). However, the results from this 

study are not limited to on-computer activities. Thus, these findings lead to the question; why 

might the very availability of computers foster more overall social interactions?  

A potentially novel way to think about this question, particularly as it pertains to the 

effects of computers in the preschool classroom, comes from a social historian’s theoretical 

perspective. This perspective suggests that it is the availability or introduction of technology 

(e.g., computers) that affects human being’s social behavior. Theorists in behavioral archeology, 

modernism and cultural critics have all consistently reported that laws of behavioral change 

cannot be explained only through nomothetic studies (Fischer, 1992; Schiffer 1992). These 

theorists recommend that we also need to study and accept that artifacts or material culture, such 

as technology, always impacts human behavior. They report that the availability of transportation 

technologies such as bicycles and cars, household technologies such as vacuum cleaners and 

refrigerator and communication technologies, such as radio, telephones, televisions, and 

computers has impacted social behaviors (Fischer, 1992; Hughes Jr. & Hans, 2001). Television 

watching, for instance, has now commonly been known to replace other activities, especially 

interpersonal and social ones (Robinson & Godbey, 1997). However, technologies like email, the 

Internet and social media are still under debate (e.g., Franzen, 2000; Hampton & Wellman, 2000; 

Nie, 2001; O’Keeffe et al., 2011). Given this theoretical perspectives, it should not come as a 

surprise that computer availability in this study contributed to the presence of group differences 

on social behaviors. However, the underlying processes that contribute to this change are 

uncertain.  

A child development, empirical perspective, may also shed some light on this question. 

The computer may be a unique tool that fosters social learning. The presence of the computer 

could alter children’s views about school or themselves. Haugland (1992) found that the self-

esteem of 4-year-old children in classrooms with computers was significantly higher than the 

self-esteem of children in classrooms without computers. Haugland (1996) hypothesized that 

self-esteem increases in computer classrooms because children view computers as “adult 

machines” and when given the opportunity to explore them, children feel important, capable, and 

competent. This type of increase in self-esteem could conceptually lead to more positive social 

interactions. 

 It is also possible that the increased interaction by students could be due to reactions of 

their teachers. Teachers could be responding to the presence of the computers by altering the way 

they interact with children in some way (Zhoa & Frank, 2003). Hypothetically, availability of 

computers could influence instructional style, lead to increased monitoring, enhanced 

involvement in off-computer play, and conflict resolution. 

Computers may offer an opportunity for children to engage in an activity that promotes 

child-child interactions. Research shows that children prefer to work on computers with peers 

(e.g., Swigger & Swigger, 1984). The indirect path could be such that the nature of social 

interactions on the computer actually promotes the learning of pro-social behaviors, which 

carries over to other classroom activities. This hypothesis is consistent with Vygotsky’s 

developmental view on socially mediated learning (Vygotsky, 1978). Certainly the relation 

between the presence of computers and positive interactions is likely to be mediated by changes 



30    JANISSE ET AL.  

 
 

in other variables. Examining the ways the ecology of the classroom changes as a result of 

computer presence, or use, is a needed addition to the literature. 

 

 

Potential Limitations & Strengths 
 

The current study had limitations which should be noted. First, it was not possible to randomly 

assign teachers or students to classrooms. Consequently, effects of teacher and classroom are 

confounded with the effect of computer availability. Although every effort was made to best 

address this limitation by ensuring that the control classroom was similar to the intervention 

classroom in many ways, attributing differences between classroom conditions to effects of the 

presence of computers is provisional. Second, a larger sample of classrooms would have helped 

control for the lack of random assignment. It would have also allowed for enhanced use of HLM 

techniques, such as the ability to test and control for level-3 variables (i.e., classroom 

characteristics).  

Despite these limitations, our results are provocative in suggesting that availability of 

computers in the Head Start classroom may increase the sociability of children during free play. 

Moreover, a significant strength of this study was that it had 4300 live observations of 66 

children in four classrooms, done over eight months, which is considered sizable for an 

observational methodology. These data provided a rich, longitudinal picture of social behavior 

for these preschoolers. Furthermore, the current study attempted to address several limitations of 

previous research in this area. The examination of naturally occurring free-play activity and not 

researcher prescribed activity, the use of a control group design, extensive observations of low 

income urban children in multiple Head Start classrooms, and the use of an operationalized 

coding system are all strengths of the current study.  

 

 

Implications: Future Research & Practice 
 

The current study presents several possibilities for future research. Future studies should focus 

on possible mediators of an effect of computer availability on social interactions. For instance, 

observation and measurement of the on-computer interactions, teaching styles on and off 

computers, child-peer-adult triadic interactions on computers, gender effects, and other 

ecological variables could lead to a better understanding of how the ever-increasing availability 

of computers impacts young children and why positive effects might occur.  

Overall, the results from this study have practical implications; namely that investing 

resources into a computer center in the Head Start classroom can have a positive effect, not just 

on the cognitive development of young children, which is often the focus of computer research 

(e.g., Li & Atkins, 2004), but also on social development. The results from this study suggest 

that the positive effect may be carried over to off-computer interactions between children, which 

in itself is a notable finding. Furthermore, these findings add to the limited research that has been 

done on the effects of computers in Head Start classrooms. Thus it expands our knowledge of 

potential benefits of the computer center in early childhood education. The NAEYC and Fred 

Rogers Centers highlight the importance of using technology and media appropriately in early 

childhood settings (2012). They emphasize that technology should be used for activities that are 

educationally sound and developmentally appropriate. They also highlight the importance of 



COMPUTERS AND SOCIAL INTERACTION     31 

 

training and digital literacy for our early childhood educators. Consistent with these 

recommendations, we believe that the practical implications reported here could be further 

enhanced with greater teacher/adult involvement in the child’s computer activity, greater 

integration of the computer into classroom curriculum, and computer training and assistance for 

teachers (Bewick & Kostelnik, 2004; Bergen, 2000, 2002; NAEYC, 2012).  

 

 

REFERENCES  
 
Anderson, G. T. (1998). Comparison of the types of cooperative problem solving behavior in four learning centers: 

Computer, dramatic play, block, and manipulative. Paper presented at the National Head Start Research 

Conference, Washington, DC. 

Anderson, G. T. (2000). An empirical comparison of the proportion of cooperative play of 4-year-old preschool 

children observed as they interact in four centers: Block, computer, housekeeping, and manipulative. 

National Association for the Education of Young Children, Atlanta, Georgia. 

Baker, C. (1985). The microcomputer and the curriculum: A critique. Journal of Curriculum Studies, 17, 4. 

Barnes, B. J., & Hill, S. (1983). Should young children work with microcomputers – Logo before Lego
TM

? The 

Computing Teacher, 10(9), 11-14. 

Bergen, D. (2000). Linking technology and teaching practice (Technology in the Classroom). Childhood Education, 

76(4), 252-253. 

Bergen, D. (2002). Choosing technology to meet varied learning purposes (Technology in the Classroom). 

Childhood Education, 79(2), 114-115. 

Bergin, D. A., Ford, M. E., & Hess, R. D. (1993). Patterns of motivation and social behavior associated with 

microcomputer use of young children. Journal of Educational Psychology, 85(3), 437-445. 

Bewick, C. J. (2000). The adoption of computers as an instructional tool by Michigan Head Start teachers. 

Unpublished doctoral dissertation, Michigan State University, East Lansing, Michigan. 

Bewick, C. J., & Kostelnik, M. (2004). Educating early childhood teachers about computers. Young Children, 59(3), 

26-29. 

Bhavnagri N. P. (1987). Parents as facilitators of preschool children’s peer relationships. Unpublished doctoral 

dissertation, University of Illinois, Urbana-Champaign, Illinois. 

Borgh, K., & Dickson, W. P. (1986). Two preschoolers sharing one microcomputer: Creating prosocial behavior 

with hardware and software. In P. F. Campbell & G. G. Fein (Eds.), Young children and microcomputers 

(pp. 37-44). Englewood, NJ: Prentice Hall. 

Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models: Applications and data analysis methods. 

Newbury Park, CA: Sage Publications. 

Calvert, S. L., Rideout, V. J., Woolard, J. L., Barr, R. F., Strouse, G. A. (2005). Age, ethnicity, and socioeconomic 

patterns in early computer use: A national survey. American Behavioral Scientist, 48(5), 590-607. 

Clements, D. H. (1994). The uniqueness of the computer as a learning tool: Insights from research and practice. In J. 

L. Wright & D. D. Shade (Eds.), Young children: Active learners in a technological age (pp. 31-49). 

Washington, DC: National Association for the Education of Young Children. 

Clements, D. H., & Nastasi, B. K. (1985). Effects of computer environments on social emotional development: Logo 

and computer-assisted instruction. Computers in the Schools, 2, 11-31. 

Clements, D. H., & Nastasi, B. K. (1993). Electronic media and early childhood education. In B. Spodek (Ed.), 

Handbook of research in the education of young children (pp. 251-275). New York: Macmillan. 

Clements, D. H., & Swaminathan, S. (1995). Technology and school change: New lamps for old?. Childhood 

Education, 71, 275-281. 

Clements, D. H., & Sarama, J. (2003). Strip mining for gold: Research and policy in educational technology-A 

response to “Fool’s Gold”, Educational Technology Review, 11(1). 

Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 

37-46. 

Cordes, C., & Miller, E. (2000). Fool’s gold: A critical look at computers in childhood. Alliance for Childhood. 

Retrieved November 2, 2006 from: 

http://allianceforchildhood.net/projects/computers/computer_reports.htm. 



32    JANISSE ET AL.  

 
 
Diamond, A., & Lee, K. (2011). Interventions shown to aid executive function development in children 4 to 12 years 

old. Science, 333, 959-964. 

Donohue, W. A., Borgh, K., & Dickson, W. P. (1987). Computers in early childhood education. Journal of Research 

in Childhood Education, 2(1), 6-16. 

Elkind, D. (1996). Young children and technology: A cautionary note. Young Children, 51(6), 22-23. 

Fischer, C. S. (1992). America calling. Berkeley, CA: University of California Press. 

Floyd, K. K., Canter, L. L. S., Jeffs, T., & Judge, S. A. (2008). Assistive technology and emergent literacy for 

preschoolers: A literature review. Assistive Technology Outcomes and Benefits, 5, 92-102. 

Franzen, A. (2000). Does the internet make us lonely? European Sociological Review, 16, 427-438. 

Freeman, N. K., & Somerindyke, J. (2001). Social play at the computer: Preschoolers scaffold and support peers’ 

computer competence. Information Technology in Childhood Education, 13, 203-213. 

Hampton, K. N., & Wellman, B. (2000). Examining community in the digital neighborhood: Early results from 

Canada’s wired suburb. In T. Ishida & K. Isbister (Eds.), Digital cities: Technologies, experiences, and 

future perspectives (pp. 194-208). New York/Berlin: Springer-Verlag. 

Hartup, W. W. (1992). Having friends, making friends, and keeping friends: Relationships as educational contexts. 

Eric Digest. 

Hartup, W. W., & Brett, L. (1987). Friendship and conflict: Synergies in child development. Eric Document. 

Haugland, S. W. (1992). The effect of computer software on preschool children’s developmental gains. Journal of 

Computing in Childhood Education, 3, 15-30. 

Haugland, S. W. (1996). Enhancing children’s sense of self and community through utilizing computers. Computers 

and Young Children, 23(4), 227-230. 

Haugland, S. W. (1997). How teachers use computers in early childhood classrooms. Journal of Computing in 

Childhood Education, 8, 3-14. 

Haugland, S. W., & Wright, J. L. (1997). Young children and technology: A world of discovery. Boston: Allyn & 

Bacon. 

Heft, T. M., & Swaminathan, S. (2002). The effects of computers on the social behavior of preschoolers. Journal of 

Research in Childhood Education, 16(2), 162-174. 

Hess, R., & McGarvey, L. (1987). School-relevant effects of educational uses of microcomputers in kindergarten 

classrooms and homes. Journal of Educational Computer Research, 3, 269-287. 

Hohmann, C. (1994). Staff development practices for integrating technology in early education programs. In J. L. 

Wright & D. D. Shade (Eds.), Young children: Active learners in a technological age (p. 104). Washington, 

DC: NAEYC. 

Hughes Jr., R., & Hans, J. D. (2001). Computers, the internet, and families: A review of the role new technologies 

plays in family life. Journal of Family Issues, 22(6), 778-792. 

Judge, S., Puckett, K., & Bell, S. M. (2006). Closing the digital divide: Update from the early childhood longitudinal 

study. The Journal of Educational Research, 100(1), 52-60. 

King, J., & Alloway, N. (1992). Preschooler’s use of microcomputers and input devices. Educational Computing 

Research, 8(4), 451-468. 

Ladd, W. G. (1981). Effectiveness of a social learning method for enhancing children’s social interactions and peer 

acceptance. Child Development, 52, 171-178. 

Lepper, M. R., & Gurtner, J. (1989). Children and computers: Approaching the twenty-first century. American 

Psychologist, 44(2), 170-178. 

Li, X., & Atkins, M. S. (2004). Early childhood computer experience and cognitive and motor development. 

Pediatrics, 113(6), 1715-1722. 

Li, X., Atkins, M. S., & Stanton, B. (2006). Effects of home and school computer use on school readiness and 

cognitive development among Head Start children: A randomized controlled pilot trial. Merrill-Palmer 

Quarterly, 52(2), 69-93. 

Lipinski, J. M., Nida, R. E., Shade, D. D., & Watson, J. A. (1986). The effects of microcomputers on young children: 

An examination of free-play choices, sex differences, and social interactions. Journal of Educational 

Computer Research, 2(2), 147-168. 

Lonigan, C. J., Allan, N. P., & Lerner, M. D. (2011). Assessment of Preschool Early Literacy Skills: Linking 

Children's Educational Needs with Empirically Supported Instructional Activities. Psychology in the 

Schools, 48(5), 488-501. 

McCormick, L. (1987). Comparison of the effects of a microcomputer activity and toy play on social and 

communication behaviors of young children. Journal of the Division for Early Childhood, 11(3), 195-205. 



COMPUTERS AND SOCIAL INTERACTION     33 

 

McMillan Culp, K., Honey, M., & Mandinach, E. (2003). A retrospective on twenty years of education technology 

policy. U.S. Department of Education. Washington, DC. Available online at 

http://www.ed.gov/rschstat/eval/tech/20years.pdf 

Muller, A. A., & Perlmutter, M. (1985) Preschool children’s problem solving interactions at computers and jigsaw 

puzzles. Journal of Applied Developmental Psychology, 6, 173-186. 

National Association for the Education of Young Children. (1996). NAEYC position statement: Technology and 

young children – ages three through eight. Young Children, 51(6), 11-16. 

National Association for the Education of Young Children. (2012). A joint position statement of the NAEYC and 

the Fred Rogers Center for Early Learning and Children’s Media at Saint Vincent College: Technology and 

Interactive Media as Tools in Early Childhood Programs Serving Children from Birth through Age 8. 

 http://www.naeyc.org/files/naeyc/file/positions/PS_technology_WEB2.pdf 

National Center for Education Statistics. (2003). Young children’s access to computers in the home and at school in 

1999 and 2000 (NCES 2003-036). Washington, DC: US Department of Education. 

Nie, N. H. (2001). Sociability, interpersonal relations, and the internet: Reconciling conflicting findings. American 

Behavioral Scientist, 45(3), 420-435. 

O’Keeffe, G. S., Clarke-Pearson, K., Council on Communications and Media (2011). The impact of social media on 

children, adolescents, and families. Pediatrics, 127(4), 800-804. 

Osborne, Jason W. (2000). Advantages of hierarchical linear modeling. Practical Assessment, Research & 

Evaluation, 7(1). 

Parsad, B., & Jones, J. (2005). Internet Access in U.S. Public Schools and Classrooms: 1994-2003 (NCES 2005-

015). US Department of Education, Washington, DC: National Center for Education Statistics. 

Plowman, L., & Steven, C. (2005). Children, play, and computers in pre-school education. British Journal of 

Educational Technology, 36(2) 145-157. 

Ramsey, P. (1995). Changing social dynamics in early childhood classrooms. Child Development, 66(3), 764-773. 

Rathbun, A. H., West, J., & Hausken, E. G. (2003). Young children’s access to computers in the home and at school 

in 1999 and 2000 (NCES 2003-036). US Department of Education, Washington, DC: National Center for 

Educational Statistics. 

Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (2
nd

 

ed.). Thousand Oaks, CA: Sage Publications. 

Raudenbush, S. W., Bryk, A. S., Cheong, Y. F., & Congdon, R. (2004). HLM 6: Hierarchical and nonlinear 

modeling. Lincolnwood, IL: Scientific Software International. 

Robinson, J., & Godbey, G. (1997). Time for life: The surprising ways Americans use their time (2
nd

 ed.). 

University Park: Pennsylvania State University Press. 

Rubin, K. H., & Krasnor, L. R. (1980). Changes in the play behaviours of preschoolers: A short-term longitudinal 

investigation. Canadian Journal of Behavioural Science, 12(3), 278-282. 

Rhee, M., & Bhavnagri, N. (1991). Four year old children's peer interactions when playing with a computer. (ERIC 

Document ED 342 466).  

Scoter, J. V., Ellis, D., & Railsback, J. (2001). Technology in early childhood education: Finding the balance. 

Northwest Regional Educational Laboratories. Portland, OR. 

Shade, D. D., & Watson, J. A. (1990). Computers in early education: Issues put to rest, theoretical links to sound 

practice, and the potential contribution of the microworlds. Journal of Educational Computer Research, 

6(4), 375-392. 

Shahramin, M. I., & Butterworth, D. M. (2002). Young children’s collaborative interactions in a multimedia 

computer environment. Internet and Higher Education, 4, 203-215. 

Schiffer, M. B. (1992). Technological perspectives on behavioral change. Tuscon: University of Arizona Press. 

Smith, P. K. (1973). Temporal clusters and individual differences in the behavior of preschool children. In R. P. 

Michael & J. H. Crook (Eds.), Comparative ecology and behaviour of primates. New York and London: 

Academic Press. 

Smith, P. K. (1978). A longitudinal study of social participation in preschool children: Solitary and parallel play 

reexamined. Developmental Psychology, 14(5), 517-523. 

Svensson, A. (2000). Computers in school: Socially isolating or a tool to promote collaboration. Journal of 

Educational Computing Research, 22(4), 437-453. 

Swigger, K. M., & Swigger, B. K. (1984). Social patterns and computer use among preschool children. AEDS 

Journal, 35-41. 

Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5
th

 ed.). Pearson Education, Inc./Allyn and 

Bacon. 

http://www.ed.gov/rschstat/eval/tech/20years.pdf
http://www.naeyc.org/files/naeyc/file/positions/PS_technology_WEB2.pdf
http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2005015
http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2005015


34    JANISSE ET AL.  

 
 
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: 

Harvard University Press. 

Zhao, Y., & Frank, K. A. (2003). Factors affecting technology uses in schools: An ecological perspective. American 

Educational Research Journal, 40(4), 807-840.