Analysis of two early childhood education settings: classroom variables and peer verbal interaction.
Article Type:
Classroom environment (Research)
Early childhood education (Psychological aspects)
Early childhood education (Models)
Early childhood education (Social aspects)
Activities of daily living (Research)
Teachers (Practice)
Teachers (Influence)
Children (Behavior)
Children (Educational aspects)
Hojnoski, Robin L.
Margulies, Allison S.
Barry, Amberly
Bose-Deakins, Jillaynne
Sumara, Kimberly M.
Harman, Jennifer L.
Pub Date:
Name: Journal of Research in Childhood Education Publisher: Association for Childhood Education International Audience: Academic Format: Magazine/Journal Subject: Education Copyright: COPYRIGHT 2008 Association for Childhood Education International ISSN: 0256-8543
Date: Winter, 2008 Source Volume: 23 Source Issue: 2
Event Code: 290 Public affairs; 310 Science & research; 200 Management dynamics Canadian Subject Form: Child behaviour
Geographic Scope: United States Geographic Code: 1USA United States

Accession Number:
Full Text:
Abstract. Descriptive and ecobehavioral analyses were used to explore the daily activity contexts in classroom settings reflecting two distinct models of early childhood education. Activity context, social configurations, teacher behavior, and child behavior were explored, with specific consideration given to peer verbal behavior as an indicator of social interaction. Twenty-four children between the ages of 3 and 6 years enrolled in a Montessori classroom and 26 children between the ages of 3 and 5 years enrolled in a traditional preschool classroom were observed over a 3-month period using the Ecobehavioral System for Complex Assessment of Preschool Environments (ESCAPE; Carta, Greenwood, & Atwater, 1986). Overall, activity context, social configurations, teacher behavior, and child behavior varied across settings in ways consistent with program philosophies. However, levels of peer verbal interaction did not vary significantly.


Increasing numbers of children are spending at least part of their day in alternate care and learning environments. In 2005, 57% of children ages 3-5 attended center-based programs alone (U.S. Department of Education, National Center for Education Statistics [NCES], 2007). The range of programs in which children may be enrolled is broad, and so, too, are the guiding philosophies and educational goals behind the programs (Barnett, 1986). Whatever the model, classroom environments for young children influence children's behavior (Odom, Favazza, Brown, & Horn, 2000). In each classroom, variable exist that can be grouped into physical, programmatic, and social environments (Nordquist & Twardosz, 1990). Variables in each of these categories may affect children's behavior. For example, research has demonstrated differences in children's social behavior under different environmental conditions, including type of toys available, group density, and length of activity periods (Brown, Fox, & Brady, 1987; Quilitch & Risley, 1973; Tegano & Burdette, 1991). To understand children's behavior and the impact of the environment, an examination of the ways in which children interact with their ecology is required.

Ecobehavioral analysis is an observational approach that focuses on the relation between child behavior and ecological variables (Greenwood, Carta, Kamps, & Arreaga-Mayer, 1990). As opposed to examining the global quality of an early childhood program and how it might affect child behavior (Dunn, 1993; Howes & Smith, 1995), ecobehavioral approaches address the interaction of people and ecological variables at a micro-level. Specifically, ecobehavioral approaches examine the extent to which an individual's behavior covaries with aspects of an individual's ecology, including materials, activities, and other individuals in the environment (Kontos, Burchinal, Howes, Wisseh, & Galinsky, 2002).

Previous applications of ecobehavioral analysis in early childhood have focused primarily on special education and regular education settings and how differences in classroom variables across those settings impact child behavior (Brown, Odom, Li, & Zercher, 1999; Kontos, Moore, & Giorgetti, 1998; McCormick, Noonan, & Heck, 1998). For example, Odom, Peterson, McConnell, and Ostrosky (1990) observed the classroom ecologies of special education preschool classrooms and early education classrooms to identify those activities most likely to encourage peer interaction. Results indicated that children in special education classrooms spent a substantial amount of time in gross motor, pre-academic, and snack activities, whereas children in early education classrooms spent more time in play, class business, and clean-up activities. In terms of social interaction, which was defined as peer verbal interaction, the base rate for children in early education classrooms was significantly higher than the base rate for children in special education classrooms. Additionally, conditional probabilities indicated that peer verbal interaction occurred significantly above the base rate for both groups during play activities.

Analysis of child and classroom characteristics among typically developing children is less common (Kontos et al., 2002; Kontos & Keyes, 1999). Given the range of philosophies and the resulting curriculum and program models, it is possible that child behavior varies greatly in settings that reflect differences in program philosophies. However, the covariation of target child behaviors and environmental variables in early childhood settings with distinctly different guiding philosophies does not appear to have been documented. Thus, the purpose of this study was to conduct a descriptive and ecobehavioral analysis of two distinctly different models of early childhood education: a Montessori program and a traditional program. These programs were specifically chosen because of the distinct way in which the classroom environment and activities reflect different guiding philosophies. For example, Montessori classrooms typically are equipped with specific materials (e.g., letter and sentence cards, cylinders of varying lengths, counting rods) that are very different from what might be used in a more traditional classroom (e.g., blocks, dramatic play materials). Further, there is a large emphasis on reality in the Montessori curriculum, whereas traditional classrooms typically include an area devoted to dramatic play and cultivating imagination (Humphryes, 1998). Given research that suggests that social interaction and complex interactions are more likely to occur during pretend or dramatic play (Kontos et al., 2002; Kontos & Keyes, 1999; Odom et al., 1990), it was expected that such differences in classroom environments would impact children's social behavior. Specifically, it was hypothesized that greater levels of peer verbal interaction would be found in the traditional setting when compared to the Montessori setting.

Given the small number of participants, this study is viewed as a pilot, demonstrating the use of a descriptive method for examining how differences in program philosophy may co-vary with contextual factors, and with teacher and child behaviors. Descriptive analyses are exploratory in nature and are intended to describe, without evaluation, the classroom ecologies of different models of early education. Ecobehavioral analyses are intended to explore the co-variation of specific classroom variables with verbal peer interaction, an indicator of peer social interaction consistent with Odom et al. (1990). In discussing results, descriptive comparisons between the two settings are made; no attempt was made, nor was it the intention of the authors, to examine the differences between the two settings from a statistical standpoint. This study extends previous research by focusing on the relation between child behavior and classroom characteristics among typically developing children in early childhood education settings with different guiding philosophies.


Participants and Settings

Fifty children enrolled in two early childhood programs participated in the study. Twenty-four children were enrolled in a Montessori program and 26 children were enrolled in a traditional early childhood education program. None was identified as having a disability, and English was the primary language of all but two participants. In the Montessori program, children ranged in age from 3 years, 1 month to 6 years, 2 months; 11 participants were boys (45.8%) and 13 participants were girls (54.2%). In the traditional program, children ranged in age from 3 years, 6 months to 6 years, 6 months; 16 participants were boys (61.5%) and 10 participants were girls (38.5%). Both programs served predominantly white children of higher socioeconomic status, according to the report of the program director; no data were collected on these variables. A letter of invitation explaining the study was sent to the parents or legal guardians of the children enrolled in each program. Only those children for whom consent was obtained were included in observations. Although differences between participants and non-participants were not formally evaluated, the homogeneity of both programs suggests there would be few differences between the two groups in terms of typical demographic variables.

Both settings were located in a large city in southwest Tennessee, and each was selected to represent a specific early childhood education philosophy. The Montessori program was selected because it was affiliated with and certified by the Association Montessori International (AMI), and the traditional program was selected as a representative of an early childhood education program accredited by the National Association for the Education of Young Children (NAEYC). Each program employed a curriculum and approach to early childhood education that was substantially different from one another, and each setting adhered to the specific philosophy that guided its practices, according to program directors. Additionally, each setting was recommended as exemplary by an early childhood program director affiliated with the local university. It should be noted that the defining characteristics of the Montessori program and the traditional programs participating in this study may not reflect potential differences that may be found regionally in each type of program, nor do they account for the existence of programs that are affiliated with both Montessori and NAEYC accreditations, and thus, have a more eclectic approach to early education.

Montessori Setting. The Montessori setting consisted of one classroom of children between the ages of 3 and 6 years. Mixed-age groupings were used so that children become accustomed to interacting with both older and younger children and so that the older children could serve as both models and teachers to their younger peers (Humphryes, 1998; Patterson, 1977). The setting was divided by bookcases into areas consistent with the Montessori curriculum (Humphryes, 1998): 1) practical life (e.g., care of self and the environment); 2) sensorial (e.g., discriminatory use of senses); 3) mathematics (e.g., concrete materials for basic concepts and computations); 4) culture (e.g., materials for organizing the biological and physical world, coordination of music, movement, and song); and 5) language arts (e.g., phonetic program of individual sounds in language). Each of the areas contained Montessori learning materials, including polishing materials, an iron and ironing board, sensorial objects, number roads, and letter and sentence cards. Aesthetically, the setting reflected the Montessori emphasis on simplicity and minimal distraction (Humphryes, 1998); with the exception of a few paintings, there were no commercial materials, posters, or bulletin boards.

All observations were conducted during the morning activity time, which is an uninterrupted three-hour work cycle characteristic of Montessori programs (Humphryes, 1998). During observations, there were approximately 30 children and three adults present, and children moved about the room independently, selecting and engaging with materials, generally with little direction or assistance from the classroom teachers. Because children are encouraged to choose activities and materials independently, children were not engaged in the same activities at the same time (Humphryes, 1998; Patterson, 1977). Consistent with Montessori philosophy, the teachers in the Montessori setting provided assistance and instruction when needed throughout the morning to individual children and to small groups of children engaged in a variety of activities, acting as facilitators of learning (Humphryes, 1998).

Traditional Setting. The second setting was a religiously affiliated early childhood education program in which children were grouped in classrooms according to age. To include the same age range as the Montessori program, a total of five classrooms were observed during morning activities: two classrooms of 3-year-old children, one classroom of 4-year-old children, and two classrooms of 5-year-old children. Each classroom was equipped with multiple activity centers designated by tables or divided by bookshelves, and including a snack table, a reading center, an arts and crafts table, a dramatic play area, and a learning table for math or writing. The setting was decorated with brightly colored posters and bulletin boards with commercial materials and children's work. There were between five and 15 children and two teachers in each classroom during observations.

The traditional setting reflected NAEYC standards for developmentally appropriate practice in that classrooms incorporated time and materials for play, child-directed learning activities, and large- and small-group activities (NAEYC, 1997). Teachers conducted large-group activities, such as circle time, to assign class jobs and discuss class business (e.g., discussing the date and weather). Teachers introduced the schedule and new center activities before the children engaged in free-play activities or other scheduled large-group activities, such as music or playground time. The emphasis on play in the traditional setting is characteristic of the principles of child development and learning that are reflected in developmentally appropriate practices (NAEYC, 1997). The teachers' roles throughout the day appeared to be direct; they guided activities, such as reading stories, during the morning routine, leading art and writing projects during free-play activities, and assigning children to centers for free-play. The teachers also monitored classroom behavior and prompted the children about expected behaviors, such as using "inside voices" and treating each other with respect.

Data Collection

Observations were conducted using the computerized Ecobehavioral System for Complex Assessment of Preschool Environments (ESCAPE; Carta, Greenwood, & Atwater, 1986) with the help of laptop computers. ESCAPE employs a momentary time sampling procedure that utilizes ecological variables in coding observations. A 60-second cycle is divided into four intervals, and when signaled by a tone, a designated child is observed. After each tone, observers had 15 seconds to code specific ecological variables, teacher variables, and child variables that were observed. This process is repeated for the duration of the observation session.

The broad categories of ecological, teacher, and child variables are broken down further into descriptors that are mutually exclusive. For example, during a Designated Activity Ecological Variable interval, the observer can only code one activity. Data were collected for all variables.

However, only the variables of Verbal Behavior, Designated Activity, Grouping, Teacher Behavior, and Teacher Focus were used in the molar and ecobehavioral analyses. Operational definitions for these variables are included in Table 1.

Data Collected

The first author, an undergraduate student, and five graduate students trained in observation and recording methods conducted a minimum of three 30-minute observations for each child participant over a 12-week period. Prior to data collection, observers spent two weeks at each setting so that children could become accustomed to the presence of new adults and computers in the classrooms. A total of 2,700 minutes of observations were conducted in the Montessori setting and a total of 2,790 minutes were conducted in the traditional setting. All observations were conducted during morning activity times (8:30 a.m. to 10:30 a.m.) and sampled a range of activities, such as music, outdoor play, and snack. Observations were conducted randomly in each setting, and no attempt was made to equate times of observations or activities observed beyond conducting all observations within the morning time period when most classroom activity occurred (according to the center directors).

Interobserver Agreement

Prior to beginning data collection, observers were trained in observation procedures and operational definitions for coded behaviors, using the ESCAPE training procedures. Each observer conducted three 30-minute observations and met a 90% criterion in all three sessions before collecting data for the study. Interobserver agreement data were collected by having two observers simultaneously observe the same child. Observations were analyzed by comparing the agreement and disagreement of the observers on each variable for each interval. The percentage of agreement was determined by the ESCAPE reliability program, which calculates interobserver agreement by dividing the number of agreements by the number of agreements plus disagreements and multiplying the result by 100%. The analysis yields an overall agreement percentage as well as agreement percentages for each variable and a Kappa coefficient. After data collection began, interobserver agreement checks were obtained for approximately 20% of the total observations conducted. Overall agreement percentages ranged from 83% to 98% and the overall Kappa was .859. Interobserver agreement percentages for specific variable categories are as follows: Designated Activity (83%-100%); Grouping (83%-100%); Teacher Behavior (74%-100%); Teacher Focus (58%-100%); and Verbal Behavior (74%-100%). When reliability agreements dropped below 85% in a variable category of interest or overall, areas of disagreement were scrutinized to determine sources of error. In conducting analyses, data with reliability agreements less than 85% in the variable of interest were dropped from the analyses; only 9 files totaling 270 minutes were excluded from the analyses overall, which is less than 5% of the entire sample of data.


Descriptive Analyses

Descriptive analysis of selected variables was conducted for each setting by using the Ecobehavioral Assessment Systems Software (EBASS; Greenwood, Carta, Kamps, & Delquadri, 1997) computer program. This analysis computes the percentage occurrence of selected variables. Data were generated by combining all child data files within a setting and selecting a variable of interest. Variables selected for the descriptive analysis were Designated Activity, Grouping, Teacher Behavior, Teacher Focus, and Verbal Behavior. These variables were selected to describe the activity context, social configurations, teachers' role in the classroom, and peer verbal interaction, respectively. Peer verbal interaction was the initial variable of interest; the other variables were selected based on research suggesting that type of activity, social configurations, and teacher behavior affect peer social interaction. Results are presented as proportion of intervals observed.

Activity Context

Given research that suggests children engage in more peer interaction during play (Kontos et al., 2002; Kontos & Keyes, 1999; Odom et al., 1990), children's activities in each program were analyzed descriptively. Table 2 presents the results of this analysis. The most visible difference between the settings is in the proportion of intervals observed in Pre-Academic and Play activities. In the Montessori program, children spent the highest proportion of time engaged in Pre-Academic activities, whereas children in the traditional program spent the highest proportion of time engaged in Play activities. Additionally, children in the Montessori program were observed in Transition activities more than children in the traditional program. In other designated activities, proportion of intervals observed was similar in both settings.

Because the Montessori program used mixed-age grouping and the traditional program grouped children by age, a descriptive analysis of Designated Activity was conducted separately by age of the child. Because of the small number of 6-year-old children, 5- and 6-year-old children were collapsed into one group. Results by age were consistent with the overall results from each program, suggesting little difference in activity context in each setting when age was considered.


Social configurations have been demonstrated to differentially affect children's behavior (Kontos & Keyes, 1999; Kontos & Wilcox-Herzog, 1997); thus, a descriptive analysis was conducted to examine grouping as a variable. Table 3 presents the results of this analysis. In general, children in each program engaged in different social configurations during activity time, with Solitary and Small Group arrangements predominant in the Montessori program and Small Group and Large Group arrangements predominant in the traditional program. A descriptive analysis conducted separately for each age group suggested that grouping patterns were similar in each setting when age was considered.

Teacher Behavior and Teacher Focus

Research suggests that the teacher's role in the classroom affects young children's behavior (Chandler, Fowler, & Lubeck, 1992; Kontos et al., 2002; Kontos & Wilcox-Herzog, 1997); thus, descriptive analyses were conducted examining Teacher Behavior and Teacher Focus to further describe the role of the teacher in the classroom. Table 4 presents the results of these analyses. In both settings, No Response (e.g., the teacher is not engaged in an observable verbal or physical response directed at any child) was coded for the greatest proportion of the intervals observed. In the Montessori program, the only other coded category accounting for more than .10 of the total behavior observed was Verbal Prompt, which includes asking questions or giving directions to children. In the traditional program, Verbal Prompt, Discussion, and Verbal Instruction each accounted for more than .10 of the total behavior observed.

In terms of Teacher Focus, in both settings, behaviors directed toward children other than the target child were coded for the largest proportion of the observations, with similar proportions across the two settings. In the Montessori program, however, the None category accounted for a large proportion of time observed, indicating the teacher was not focused on any child in particular, whereas in the traditional program, Target and Other accounted for the second largest proportion of time observed.

Verbal Behavior

A descriptive analysis of Verbal Behavior was conducted to examine Talk to Peer as an indicator of the level of peer social interaction. The overall results are depicted in Table 5. In general, in both settings, little verbal interaction was coded, either with peers or teachers. To further examine peer social interaction, a descriptive analysis of Talk to Peer was conducted by age. As Table 6 demonstrates, in the Montessori program, differences between the age groups appear small, whereas in the traditional program, differences between the age groups are more visible.

Summary of Descriptive Analyses

Overall, results from the descriptive analyses suggest that classroom ecologies reflect the guiding philosophies of the programs. Specifically, activity context in the programs differed in terms of the emphasis on the work cycle of the Montessori program and play in the traditional program. Results related to grouping were consistent with the structure evidenced in the routine and schedule of the programs, with children in the Montessori program spending the larger proportion of time engaged in Solitary activities and children in the traditional program spending the larger proportion of time engaged in Large-Group activities. Somewhat counterintuitively, in both settings, No Response was the predominant code for teacher behavior and No Talk was the predominant code for child Verbal Behavior. Several interpretations of these findings in the context of the study are addressed in the discussion.

Ecobehavioral Analyses

Although descriptive analyses provide information about classroom and child variables, the way in which the classroom variables interact with child variables is not evident; a more dynamic analysis is needed. The EBASS computer system was used to conduct an ecobehavioral analysis of peer verbal interaction, as it occurred under different ecological and teacher conditions. This analysis computes the probability of child behavior, as it occurs simultaneously with the ecological variables. The base probability, or unconditional probability, of the behavior is compared with the occurrence of the behavior under selected conditions to determine if selected conditions are associated with increases or decreases in the probability of the behavior. For the purpose of this study, Talk to Peer was selected as the target behavior, or general indicator of social interaction, consistent with the findings of Odom et al. (1990). Designated Activity, Grouping, Teacher Behavior, and Teacher Focus were selected as the ecological and teacher conditions, based on research suggesting these variables may impact social interaction. Only activities occurring more than .05 of the observed time were included in the analysis, as these are thought to provide a stable estimate of the ecobehavioral relation (Greenwood, Delquadri, Stanley, Terry, & Hall, 1985).

Peer Verbal Interaction in Designated Activities

In the Montessori program, none of the Designated Activities was statistically significant in terms of their relation to Talk to Peer. That is, Talk to Peer was not likely to occur either above or below the base rate level of .15 during any Designated Activity. In the traditional program, the base rate of Talk to Peer was .17, and Play as a Designated Activity increased the probability that children engaged in Talk to Peer above the base rate (.33), as did Snack (.22), with both results significant at the p < .001 level. Designated Activities that were associated with probabilities of Talk to Peer significantly less than the base rate included Transition (.15), Fine Motor (.11), Pre-Academic (.07), Class Business (.04), and Story (.02), all of which were significant at least at the p < .05 level.

Peer Verbal Interaction and Grouping

Given differences in the Grouping in the two settings and previous social configuration research (Kontos & Keyes, 1999; Kontos & Wilcox-Herzog, 1997), an ecobehavioral analysis of Grouping and Talk to Peer was conducted to explore the relation between the two variables. In the Montessori program, the overall probability of Talk to Peer increased significantly above the base rate of. 15 when the children were observed in Small Group (.24) and decreased significantly below the base rate when children were observed as Solitary (.06), with both results statistically significant at the p < .001 level. Examination of the relation between grouping and peer verbal interaction by age reflected a similar pattern. In the traditional program, overall, the analysis indicated a statistically significant increase above the base rate of .17 in the probability of Talk to Peer in Small Group (.28) and a statistically significant decrease in the probability of Talk to Peer in Large Group (.07), with both results significant at the p <. 001 level. Examination by age reflected a similar pattern.

Peer Verbal Interaction and Teacher Variables

In light of previous findings indicating that children engage in less complex interactions with their peers when teachers are involved in the activity (Kontos et al., 2002), an ecobehavioral analysis was conducted using Teacher Behavior, Teacher Focus, and Talk to Peer to examine the effects of teacher presence on children's social interactions. In the Montessori program, no significant results were found when examining the occurrence of Talk to Peer under different conditions of Teacher Behavior or Teacher Focus. That is, Talk to Peer was not likely to occur at either above or below the base rate level of .15 under differing conditions of Teacher Behavior or Teacher Focus. The traditional program showed statistically significant increases at the p < .05 level in the probability of Talk to Peer occurring above the base rate of .17 when the Teacher Behavior was No Response and the Teacher Focus was None. The probability of Talk to Peer increased further when Teacher Focus was Other (p < .001). There was a statistically significant decrease at the p < .001 level in Talk to Peer when the Teacher Behavior was Read/Sing (.04) and when Teacher Focus was Target and Other (.07) and Target (.04).

Summary of Ecobehavioral Analyses

In general, findings from the traditional program reflect previous research examining the effect of classroom variables on peer social interaction. Specifically, Talk to Peer was more likely to occur during play activities and in the absence of teacher presence. Findings from the Montessori program, however, appear to be unique, albeit difficult to interpret. Ecobehavioral analyses did not yield activity contexts or teacher variables that were associated with increases or decreases in peer social interaction, as indicated by Talk to Peer. Equally interesting is the analysis of grouping and peer verbal interaction; in both settings, Talk to Peer was less likely to occur in the social configuration in which children spent the largest proportion of their time, according to the descriptive analysis. That is, in the Montessori program, Solitary decreased the probability of Talk to Peer, and in the traditional program, Large Group decreased the probability of Talk to Peer.


In general, the results of the study indicate that children in the two models of early childhood education were observed in activity contexts consistent with and reflective of the distinct program philosophies. That is, children in the Montessori program were observed primarily in Pre-Academic activities, whereas children in the traditional program were observed primarily in Play activities. Although differences in activity context between the two settings may seem logical and predictable, given the guiding program philosophies, this study documents those differences and supports the notion that environmental differences impact children's behavior, or the activities in which children are engaged. Further, in general, the patterns of activity context were consistent across children of different ages, suggestive of the influence of the environment beyond simple developmental differences. This difference in activity context is particularly important, given research (Cemore & Herwig, 2005; Elias & Berk, 2002; Gayler & Evans, 2001; Krafft & Berk, 1998), developmental theory (e.g., Lev Vygotsky and Jean Piaget), and position statements (NAEYC, 1997) that emphasize the importance of play in development.

Different educational philosophies also seemed to be reflected in the differences in social configurations. That is, although both groups of children spent similar amounts of time in small groups, children in the Montessori program were observed spending a large proportion of time in solitary activities and children in the traditional program spent a large proportion of time in large-group activities. These social configurations have the potential to affect social behavior, as indicated by the ecobehavioral analysis of Grouping and Talk to Peer. Small groups in both settings were associated with greater occurrences of Talk to Peer, and solitary and large group arrangements were associated with lower probabilities in the Montessori and in the traditional settings, respectively. In both settings, it appears that small groups were more conducive to peer verbal interaction, which suggests that this social configuration could be used strategically to promote such interaction. It is likely that in the large group, either peer verbal interaction is less acceptable, as is the case when it is a teacher-directed activity, or less frequent because of the increased number of participants, as might be the case with a large group of children engaged in play.

Results of the analysis of the occurrence of Verbal Behavior are particularly interesting for several reasons. Given the differences in activity contexts in the two settings (specifically, the difference in levels of Play), differences in grouping, and earlier research indicating that children engage in peer interaction more frequently during play (Kontos et al., 2002; Kontos & Keyes, 1999; Odom et al., 1990), one might have predicted the occurrence of Talk to Peer to be higher for children in the traditional program, consistent with the hypothesis of the study prior to data collection. However, ecological differences in the settings did not result in a large discrepancy in peer interaction as represented by Talk to Peer; children in both settings had similar base rates of Talk to Peer. Thus, the hypothesis was not supported.

One possible explanation for the finding is that differences in the environment resulted in qualitative differences in verbal behavior as opposed to differences in the pure frequency of the behavior. For example, anecdotal observations in the Montessori program indicated that Talk to Peer occurred more as a means of communicating about Pre-Academic material or activities (e.g., inviting a peer to work with them and talking about materials). In the traditional program, Talk to Peer appeared to occur not only as a means of communicating about materials, but also as a way to facilitate dramatic and pretend play, a function that appeared to be absent in the Montessori program during observations.

Although no specific data were collected to address the potential function of Talk to Peer, qualitative differences between the two settings are reflected in the ecobehavioral analysis of Talk to Peer and Designated Activity and Talk to Peer and Teacher Behaviors, partially supporting the function hypothesis. In the Montessori program, the likelihood of Talk to Peer occurring was similar across all activities, with none statistically significant. The occurrence of Talk to Peer across activities and in the virtual absence of Play suggest that children may have used peer verbal interaction in a different way than might occur during Play. In the traditional program, the likelihood of Talk to Peer occurring was greatest when children were engaged in Play and Snack. Play and Snack tend to be more social as opposed to academic activities; thus, it is possible that Talk to Peer served as a means of social interaction. Conversely, the probability of Talk to Peer occurring was lowest for Story, Class Business, and Pre-Academic, in that order. These activities tend to be teacher-directed and peer verbal interaction would not be encouraged, or perhaps even acceptable. The ecobehavioral analysis of Talk to Peer and Teacher Behaviors support this notion, in that there was a decreased probability of Talk to Peer when Teacher Behavior was coded as Read/Sing and Teacher Focus was Target and Other or Target.

In addition to minimal differences in the base rates of Talk to Peer, in both settings, children spent the largest proportion of time not engaged in verbal behavior. Although the base rate for Talk to Peer was higher in both settings than that found by Odom et al. (1990), the large proportion of time spent with no verbal behavior is of interest. One explanation is that children may have been engrossed in their activities, whether it be Pre-Academic or Play. Another explanation is that the result may be reflective of the developmental stage of the children, a stage at which there is less emphasis on verbal communication. Indeed, a developmental progression is reflected in the analysis of Talk to Peer by age in the traditional program, with the behavior increasing as the age of the child increases. This pattern was not evident in the Montessori program, however, and potentially may be related to the differences in grouping. It may be that Montessori's multi-age grouping facilitates verbal interaction, with the older children providing advanced models for the younger children (Humphryes, 1998). Future research may address the effect of multi-age grouping on certain child outcomes, such as language, social competence, and cognitive performance. A final explanation for the large proportion of No Talk may be the constraints of the coding system. Because ESCAPE uses a momentary time sampling method and only one child is observed at a time, it is likely that children were engaged in verbal behavior that was not captured by the coding system.

There are several limitations of the present study that are important to consider when interpreting the results. Certain artifacts of the methodology deserve consideration in terms of their effect on the results. Difficulties were encountered during observations and coding in terms of the materials used. This may have resulted in an artificial inflation of the Pre-Academic category in the Montessori program. A Montessori program uses materials that are specifically designed to reflect conceptual challenges and build on previous concepts. For example, a puzzle of the African continent may serve as a manipulative to develop fine motor skills and also serve as a tool to teach social studies and geography; these materials would be coded as Instructional, and the Designated Activity would be coded as Pre-Academic. In a traditional program, the puzzle may depict cartoon characters and, thus, primarily serve as a means of developing fine motor skills; materials would be coded as Manipulative and the Designated Activity as Fine Motor. The central consideration in coding Pre-Academic over another code was whether or not the material or activity was related to reading, mathematics, handwriting, science, social science, or health and safety. Consequently, topographically similar behavior may have been coded differently in the two different settings. Although this may require qualification of the results, this apparent difference between the two settings is noteworthy in and of itself. That is, the very nature of the materials in the Montessori setting potentially supports different behaviors than those supported by materials used in other settings. Other primary variables of interest were not subject to the same potential discrepancies in coding, as these variables were identical across settings; thus, no qualification is needed in interpreting those results. Finally, the ESCAPE system could not be used to directly compare the findings in the two settings. That is, results were calculated separately for each setting and then compared descriptively. Statistical analysis of the results would determine whether differences observed reached statistical significance.

Although Talk to Peer was not subject to coding difficulties and appears to provide a consistent representation of the same behavior across both settings, it is a narrow and limited representation of such a complex topic as social interaction. The decision to use Talk to Peer to represent social interaction, while consistent with other research, meant that other aspects of social interaction would not be included or captured in the data. For example, if the target child were engaged in social interaction during the observation, but at the precise moment of the tone signal was listening to a peer, the behavior would not be coded as Talk to Peer.

A final limitation is the small sample size and the use of only one setting as representative of each program type. This significantly limits the generalizability of the results. Although the settings were selected as ably representing the espoused program philosophies (according to indicators of accreditation and personal recommendation), a larger sample would increase the generalizability of the results. In addition, data from the traditional classrooms were aggregated, which may have obscured differences in individual classrooms. However, analyses by age for Activity Context and Grouping yielded results similar to those found for all the classrooms combined. In addition, data were aggregated to facilitate descriptive comparisons between the multi-age Montessori setting and the traditional setting. Finally, because no data were obtained regarding philosophy fidelity beyond accreditation and director report, it is unclear the extent to which each of the settings accurately represented the particular philosophies. It is possible that the behaviors and ecological variables observed were particular to the specific setting and not reflective of the overall guiding philosophy. In future studies, philosophy fidelity could be documented by videotaping classroom activities and having experts review the videotapes and code the setting, based on descriptions of various philosophies.

This study focused exclusively on a descriptive exploration of classroom ecologies and an exploration of child behavior as it covaried with ecological variables in two different early childhood education environments. No attempt was made to examine child outcome variables. However, the findings of the study have implications related to child outcomes. Developmental theory, policy, and research suggest that sociodramatic play may be important for cognitive, social, and emotional development. The absence of sociodramatic play in the Montessori program, then, would seem to be a potential disadvantage. However, it is likely there are other aspects of the Montessori program that provide opportunities for similar development, thereby compensating for the absence of sociodramatic play. It is critical to identify the features of classroom environments--Montessori or otherwise--that support young children's positive development. Differences in the types of activities provided for children and the types of activities in which they engage may result potentially in different outcomes, or differences in the level of achievement. Although Montessori education was not associated with higher levels of academic achievement in the upper grades (Lopata, Wallace, & Finn, 2005), the relation between program model and early skill development needs to be more clearly defined. Given the importance of quality early education, independent of the model, to later outcomes, the relation between early education program models and indicators of growth and development in such areas as language, early literacy, early numeracy, motor skills, social competence, and self-regulation is worthy of exploration.


Barnett, D. W. (1986). School psychology in preschool settings: A review of training and practice issues. Professional Psychology: Research and Practice, 17(1), 58-64.

Brown, W. H., Fox, J. J., & Brady, M. P. (1987). Effects of spatial density on 3- and 4-year-old children's socially directed behavior during free play: An investigation of a setting factor. Education and Treatment of Children, 10(3), 247-258.

Brown, W. H., Odom, S. L., Li, S., & Zercher, C. (1999) Ecobehavioral assessment in early childhood programs: A portrait of preschool inclusion. The Journal of Special Education, 33(3), 138-153.

Carta, J. J., Greenwood, C. R., Atwater, J. B. (1986). ESCAPE: Eco-behavioral system for the complex assessment of preschool environments. Unpublished coding manual for observational system. Juniper Gardens Children's Project, Kansas City, KS.

Cemore, J. J., & Herwig, J. E. (2005). Delay of gratification and make-believe play of preschoolers. Journal of Research in Childhood Education, 19(3), 251-267.

Chandler, L. K., Fowler, S. A., & Lubeck, R. C. (1992). An analysis of the effects of multiple setting events on the social behavior of preschool children with special needs. Journal of Applied Behavior Analysis, 25, 249-263.

Dunn, L. (1993). Proximal and distal features of day care quality and children's development. Early Childhood Research Quarterly, 8, 167-192.

Elias, C. L., & Berk, L. E. (2002). Self-regulation in young children: Is there a role for sociodramatic play? Early Childhood Research Quarterly, 17, 216-238.

Gayler, K. T., & Evans, I. M. (2001). Pretend play and the development of emotion regulation in preschool children. Early Child Development and Care, 166, 93-108.

Greenwood, C. R., Carta, J. J., Kamps, D., & Delquadri, J. (1997). Ecobehavioral assessment systems software (EBASS Version 3.0): Practitioner's manual. Kansas City, KS: Juniper Gardens Children's Project, University of Kansas.

Greenwood, C. R., Carta, J. J., Kamps, D., & Arreaga-Mayer, C. (1990). Ecobehavioral analysis of classroom instruction. In S. Schroeder (Ed.), Ecobehavioral analysis and developmental disabilities: The twenty-first century (pp. 33-63). New York: Springer-Verlag.

Greenwood, C. R., Delquadri, J., Stanley, S. O., Terry, B., & Hall, R.V. (1985). Assessment of eco-behavioral interaction in school settings. Behavioral Assessment, 7, 331-347.

Howes, C., & Smith, E. W. (1995). Relations among child care quality, teacher behavior, children's play activities, emotional security, and cognitive activity in child care. Early Childhood Research Quarterly, 10, 381-404.

Humphryes, J. (1998). The developmental appropriateness of high-quality Montessori programs. Young Children, 53(4), 4-16.

Kontos, S., Burchinal, M., Howes, C., Wisseh, S., & Galinsky, E. (2002). An ecobehavioral approach to examining the contextual effects of early childhood classrooms. Early Childhood Research Quarterly, 17(2), 239-258.

Kontos, S., & Keyes, L. (1999). An ecobehavioral analysis of early childhood classrooms. Early Childhood Research Quarterly, 14(1), 35-50.

Kontos, S., Moore, D., & Giorgetti, K. (1998). The ecology of inclusion. Topics in Early Childhood Special Education, 18, 38-48.

Kontos, S., & Wilcox-Herzog, A. (1997). Influences on children's competence in early childhood classrooms. Early Childhood Research Quarterly, 12, 245-262.

Krafft, K. C., & Berk, L. E. (1998). Private speech in two preschools: Significance of open-ended activities and make-believe play for verbal regulation. Early Childhood Research Quarterly, 13, 637-658.

Lopata, C., Wallace, N. V., & Finn, K.V. (2005). Comparison of academic achievement between Montessori and traditional education programs. Journal of Research in Childhood Education, 20(1), 5-14.

McCormick, L., Noonan, M. J., & Heck, R. (1998). Variables affecting engagement in inclusive preschool classrooms. Journal of Early Intervention, 21, 160-176.

National Association for the Education of Young Children. (1997). Developmentally appropriate practice in early childhood programs serving children from birth through age 8. Retrieved June 7, 2004, from

Nordquist, V. M., & Twardosz, S. (1990). Preventing behavior problems in early childhood special education classrooms through environmental organization. Education and Treatment of Children, 13(4), 274-287.

Odom, S. L., Favazza, P. C., Brown, W. H., & Horn, E. M. (2000). Approaches to understanding the ecology of early childhood environments for children with disabilities. In T. Thompson, D. Felce, & F. J. Symons (Eds.), Behavioral observation: Technology and applications in developmental disabilities (pp. 193-214). Baltimore: Brookes Publishing.

Odom, S. L., Peterson, C., McConnell, S., & Ostrosky, M. (1990). Ecobehavioral analysis of early education/specialized classroom setting and peer social interaction. Education and Treatment of Children, 13(4), 316-330.

Patterson, C. H. (1977). Foundations for a theory of instruction and educational psychology. New York: Harper & Row.

Quilitch, H. R., & Risley, T. R. (1973). The effects of play materials on social play. Journal of Applied Behavior Analysis, 6(4), 573-578.

Tegano, D. W., & Burdette, M. P. (1991). Length of activity periods and play behaviors of preschool children. Journal of Research in Childhood Education, 5(2), 93-99.

U.S. Department of Education, National Center for Education Statistics [NCES]. (2007). The condition of education 2007 (NCES 2007-064). Retrieved October 1, 2007, from http://nces. on.

Correspondence regarding this article should be addressed to Robin L. Hojnoski, Lehigh University, College of Education and Human Services, Bethlehem, PA 18015; E-mail:

Robin L. Hojnoski (1)

Allison S. Margulies (2)

Amberly Barry

Jillaynne Bose-Deakins (3)

Kimberly M. Sumara (4)

Jennifer L. Harman (5)

The University of Memphis


(1) Robin L. Hojnoski is now at Lehigh University, Bethlehem, PA

(2) Allison Margulies is now at the Kennedy Krieger Institute, Baltimore, MD.

(3) Jillayne Bose-Deakins is now with the Northern Suburban Special Education District, Chicago, IL.

(4) Kimberly Sumara is now with the Memphis City Schools, Memphis, TN.

(5) Jennifer Harman is now at the University of Florida.
Table 1
Operational Definitions of Observational Categories

Variable             Definition

Verbal Behavior      "... verbal or signed expression by the observed
                     student" (p. 63).
1. NOTALK            "... lack of verbal or signed expression by the
                     observed student" (p. 63).
2. TOTEACHER         "... verbal or signed expressions by the observed
                     student that are directed at the teacher, a
                     teaching adult, or another unspecified adult"
                     (p. 64).
3. UNDIRECTED        "... verbal or signed expressions by the observed
                     student that appear to have no specific
                     direction" (p. 64).
4. TOPEER            "... verbal or signed expressions by the observed
                     student that are directed at a peer or peers"
                     (p. 64).
5. CN'TTELL          "... target child could not be observed because
                     she/he was temporarily out of sight or out of
                     earshot" (p. 64).
Teacher Behaviors    "... defined as the behavior being emitted by the
                     defined teacher" (p. 65).
1. PHYSASST          "... providing children with physical guidance or
                     help ... only coded when the 'teacher' helps the
                     child engage in a behavior" (p. 65).
2. GEST/SIGN         "... communicating via physical responses or
                     providing a physical cue to respond" (p. 65).
3. APPROVAL          "... expressing praise, appreciation, or
                     satisfaction with the Glasswork, conduct, or
                     performance of one or more students" (p. 65).
4. DISAPPROV         "... a redirection of student behavior through the
                     expression of dissatisfaction with the behavior,
                     Glasswork, appearance, or performance of one or
                     more students" (p. 66).
5. VERBPRMPT         "... asking questions or giving direction to
                     children" (p. 66).
6. VERBINSTR         "... engaging in talk (non-questions and
                     non-commands) related to the preschool
                     curriculum" (p. 66).
7. READ/SING         "... telling a story, reciting a poem or
                     fingerplay, or performing a song with one or more
                     children" (p. 66).
8. DISCUSS           "... engaging in talk unrelated to the preschool
                     curriculum" (p. 66).
9. NORESPONS         "... when the person coded in the Teacher
                     Definition variables makes neither an overt nor
                     an observable response at either the target
                     student or any other student in the class or when
                     No Staff is coded" (p. 66).
10. CN'TTELL         "... when the teacher's behavior has not been
                     observed because she/he is temporarily out of
                     sight or because the behavior does not fit any
                     of the above designations" (p. 66).

Teacher Focus        "... defined as the direction of the behavior of
                     the adult" (p. 66) coded in the Teacher
                     Definition variables.
l. NONE              "... the adult is not directing overt behavior
                     towards any student in the classroom" (p. 67).
2. OTHER             "... no adult is directing behavior towards the
                     observed student and the defined teacher is
                     directing behavior towards a student other than
                     the observed student" (p. 67).
3. TARG+OTH          "... the adult directs behaviors toward the
                     observed child and other children" (p. 67).
4. TARGET            "... the adult directs behavior toward the
                     observed child in a one-on-one situation" (p. 67).
5. CN'TTELL          "... teacher's focus has not been observed
                     because she/he is temporarily out of sight"
                     (p. 67).

Designated           "... learning experience or focus of instruction
Activity             being provided" (p. 67).
1. SNACK             "... the preparation and/or the eating of real
                     food" (p. 68).
2. PLAY              "... activities ... in which children interact
                     with toys or pretend play materials" (p. 68).
3. TRANSIT           "... the time between activities/tasks when the
                     target child is engaged in: (1) getting materials
                     for a new activity, (2) moving to a different
                     activity or location to begin a new activity, or
                     (3) waiting for a cue to begin a new activity"
                     (p. 68).
4. PREACADEM         "... devoted to the teaching/learning of skills
                     related to reading, mathematics, handwriting,
                     science, social science, health, and safety"
                     (p. 68).
5. FINEMOTOR         "... movement of small muscles to move fingers,
                     wrists, and hands, usually involving the grasp"
                     (p. 68).
6. MUS/RECIT         "... one or more children singing, dancing,
                     listening to music, clapping and moving to music,
                     doing finger- plays, or reciting poems" (p. 68).
7. CLEANUP           "... putting away toys, instructional materials,
                     worksheets, play equipment, furniture, or food"
                     (p. 69).
8. CLASSBUS          "... group discussion periods ..." (p. 69).
9. STORY             "... the reading or telling of a tale" (p. 69).
10. SELFCARE         "... teaching or the occurrence of skills that
                     involve personal needs and hygiene ..." (p. 69).
11. GRSS             "... the teaching of movement of the large
    MOTOR            muscles in the arms, legs, and feet" (p. 69).
12. TIMEOUT          "... child has been removed from the group or from
                     participation as a disciplinary action" (p. 69).
13. LANGPROG         "... instruction of specific speech or language
                     skills" (p. 69).
14. CN'TTELL         "... the above listed activities do not apply or
                     seem to fit the situation or if two or more of
                     the definitions stated above seem to fit equally
                     as well" (p. 69).

Grouping             "... defined by the number of students who are:
                     (1) engaged in the same Designated Activity and
                     (2) in the same vicinity as the observed child"
                     (p. 72).
1. SMALLGRP          "... observed child is engaged in an activity or
                     in a free choice situation and is in a location
                     with four or fewer children" (p. 73).
2. LARGEGRP          "... observed child is engaged in an activity with
                     five or more children" (p. 73).
3. ONEONONE          "... observed child is engaged in an activity by
                     him- or herself with one adult" (p. 73).
4. SOLITARY          "... observed child is engaged in an activity by
                     him- or herself" (p. 73).
5. CN'TTELL          "... size of the group cannot be
                     determined ..." (p. 73).

Table 2
Proportion of Intervals Observed in Activity Context

Designated Activity   Montessori   Traditional

Snack                    .04           .07
Play                     .04           .27
Transition               .21           .13
Pre-academic             .47           .16
Fine Motor               .09           .09
Music                    .01           .03
Clean-up                 .03           .03
Class Business           .00           .09
Story                    .00           .07
Self-care                .05           .01
Gross Motor              .01           .01

Table 3
Proportion of Intervals Observed in Social Configurations

Grouping      Montessori   Traditional

Solitary         .44           .11
Small Group      .45           .41
Large Group      .08           .41
One-on-One       .01           .03

Table 4
Teacher Variables in Proportion of Internals Observed

                     Montessori   Traditional

                         Teacher Behavior

No Response             .55           .28
Verbal Prompt           .15           .20
Discussion              .02           .15
Verbal Instruction      .06           .12

                          Teacher Focus

Other                   .49           .46
None                    .40           .20
Target                  .02           .24

Table 5
Proportion of Intervals Observed in Verbal Interaction

Verbal Behavior              Montessori   Traditional

No Talk                         .69           .60
Talk to Peer                    .15           .17
Talk to Teacher                 .01           .04
Undirected Verbal Behavior      .08           .07

Table 6
Proportion of Internals Observed in Talk to Peer By Age

Age of Child   Montessori   Traditional

3 Year-Olds       .14           .10
4 Year-Olds       .13           .17
5 Year-Olds       .16           .20
Gale Copyright:
Copyright 2008 Gale, Cengage Learning. All rights reserved.