Relationship between classroom plan types and the degree of concentration of the children in elementary schools: A comparative study of open‐plan classrooms and conventional‐plan classrooms

Abstract

In this study, the authors collected the data of 229 classrooms in 15 municipal elementary schools in a city, and sorted them into 5 categories according to the degree of openness of the classrooms in the design of borders between classroom and hallway. First, the authors examined the influence of 5 major factors on the degree of concentration of the children during classwork. As a result, the following were found: (1) the degree of concentration of the children is influenced by the classroom plan type and the number and ratio of children in the class who require special supports, and is neither affected by teaching experience of the teacher nor the total number of the children in the class. (2) The general trend of concentration of children in full‐open classrooms does not differ significantly from that in conventional classrooms. However, in the classrooms equipped with movable partition panels for partially separating the classroom from the hallway, it is significantly lower than in the classrooms of other types. The factors of distraction in physical environments were then further analyzed and discussed.

Introduction Background and purpose

During the 1970s, open-plan classrooms emerged in elementary schools in Japan. The open-plan classroom was an outcome of the open school concept, an educational experiment at the time that attempted to reform a class-centered, undifferentiated educational system. In 1984, the Ministry of Education, Science, and Culture launched a program to fund multipurpose open spaces, prompting a rapid increase in the number of schools across Japan that provide such spaces. As part of this trend, a sizeable proportion of elementary schools started providing open-plan classrooms (Figure ).

A survey among schools in Tokyo found that open-plan classrooms were featured in 60.8% of the elementary schools built or refurbished from 1981 onward, and in 8.9% of the all elementary schools as of 2013. Rather than serving as settings for innovative experimental pedagogical practices, open-plan classrooms in Japan have functioned as regular school classrooms for a broad range of daily school education practices.

Meanwhile, educators are continuing to explore teaching approaches that respect the individuality of schoolchildren. The most notable of these is the 2007 School Education Act provided for “the implementation of special needs education,” thereby institutionalizing an environment that provides special needs support to students who require it, without taking them out of regular classes. Accordingly, public elementary schools are now expected to cater to schoolchildren with a diverse range of abilities. A national survey estimated that as much as 6.5% of schoolchildren, and 7.7% of elementary schoolchildren, “exhibit significant learning or behavioral problems despite having no intellectual disabilities.” Such schoolchildren must be accounted for when considering the learning environment in the public elementary schools of today.

Thus, it is essential to determine how the increasingly diverse classroom designs in elementary schools have influenced classroom learning from 2007 onward. We were interested in determining if certain classroom environments were conducive to focused learning. To this end, variations in the level of learner concentration, as perceived by the teachers, were evaluated with respect to the difference in the design of borders between classroom and hallway.

Existing research

For many years, efforts to reform elementary school education have been focused on instructional techniques and lesson content. The fact that environment significantly affects educational outcomes has been known for some time. For example, Louis Malaguzzi, who pioneered the development of the Reggio Emilia approach, stated, “There are three teachers of children: adults, other children, and their physical environment.” As one kind of educational environment, the open-plan classroom has been evaluated as follows.

Evaluation of open-plan classrooms in the US

Regarding the US literature, and focusing on the open school plan, Gump conducted a number of reviews of the researches on the effects of the physical environment of schools on educational outcomes. According to Gump, open-plan schools are characterized by openness, which consists of 4 elements:

1. The space is not partitioned by walls and can therefore be used flexibly;
2. Given that there are no hallways, more space is made available;
3. There are no walls or doors, meaning that learners can travel directly between activity spaces;
4. External stimuli can directly reach activity spaces.

The above characteristics are appropriate for the American concept of the open school designs, which features several activity groups that utilize different sections of a vast space. The open schools that the American researchers studied correspond with this American design.

From the outset, researchers have raised concerns about the sonic environment. A review by Gump highlighted this as a flaw in the environment of the open school; Gump noted that the problem of noise, and its inhibition of the concentration of children, was ubiquitous, and that there were reports of poor academic achievement in reading comprehension, vocabulary, and arithmetic.

The evidence seemed to suggest that each problem was attributable to the distraction of children as a result of the openness of the open-plan classroom type. With respect to this, Gump raised 2 noteworthy points.

First, the issue was not solely the sound level (decibels); but message value as well. Sounds with a message value that attract the attention of the children inhibit concentration more than equally loud sounds with no meaning. Examples include a friend talking about someone, a laugh, and a teacher scolding. This implies that the noise problem in open-plan classrooms stems not from the loudness of the noises, but the ease with which the message values are transmitted.

Second, the consequences of openness are not limited to direct effects, such as the distraction of learners. Gump stated that open-plan schools were not significantly noisier or more distracting than traditional schools, the reason being that teachers in open-plan schools take care to avoid distracting learners in other regions as they teach. What is problematic is that, in seeking to avoid distracting children in other classes, teachers refrain from activities that could produce noise or smells, and by so doing, teachers are unable to flexibly develop lessons. Consequently, educational efficacy is undermined.

The review by Gump provided valuable insights into the effectiveness of open-plan classrooms as learning environments. In our attempt to evaluate the open-plan classrooms in Japan, we focused on the problem of distraction in classrooms and how the concentration of learners may be impeded. Moreover, factors other than the noise level were included in our analysis.

Gifford, who stated that behavior is the product of an interplay between individual attributes, the sociocultural context, and the physical environment, argued that classroom designs were not simply either good or bad but that they produced nuanced effects depending on certain conditions. Solomon and Kendall argued that a given set of physical conditions can produce both positive and negative effects. As an example, they cited a survey of schoolchildren who were autonomous in their behavior but not academically oriented. From the survey, it was observed that, when compared with such children in standard classrooms, those in open-plan classrooms broke the rules less frequently but received lower scores on academic tests. The researchers then highlighted, as the most significant drawback, the inadequate training of teachers in the use of open-plan classrooms. Gump also mentioned that educational programs are often ill-suited to the physical characteristics of the classroom, and that this is problematic for educational efficacy. These findings imply that with respect to school architecture, it is necessary to evaluate the architectural aspects of schools and ensure that school architecture reflects the realities of classroom teaching.

Evaluation of open-plan classrooms in Japan

As discusses previously, there have been several empirical studies on US-style open-plan classrooms. However, open-plan classrooms in Japan differ from those in America in terms of architectural configuration and how they are used. Thus, the empirical findings and outlooks described above cannot be directly applied to the situation in Japan. Accordingly, studies in Japan have examined open-plan classrooms in terms of how they are configured and used in Japan.

One architectural study of Japanese open-plan classrooms evaluated the physical performance of classrooms. Using an acoustic scale model, Omori et al examined sound insulation strategies that may improve the sonic environment. It was found that the installation of sound absorbers along the interior wall of an open-space adjoining classrooms was less effective in terms of sound insulation than the installation of partial barriers (such as hanging partitions) between the open space and the classrooms. It was also found that a higher number of barriers resulted in greater sound insulation. This finding indicates how difficult it is to solve the problem created by opening up classrooms, namely the volume of noises from neighboring classrooms, without closing off the classrooms.

Aoki et al examined the perceptions teachers had of the sonic environment in real-world open-plan classrooms. It was reported that the sound absorption of the ceiling made it easier to hear what people were saying. Moreover, it was found that the environment could not be evaluated solely on the basis of physical performance, given that the sonic environment was dependent on how the relevant school was using the classrooms, and the attitudes of the individuals using them during lesson times. Likewise, Ueno et al and Hashimoto et al reported that the sonic environment can be improved with designs that consider physical sonic characteristics; however, it was revealed that classroom environments are evaluated in an integrative manner. For example, it was highlighted that the sonic environment is not solely determined by the sonic acoustic performance, and that elements external to the sonic environment can affect the overall evaluation. In this study, rather than assessing the performance of the physical environment, the concentration levels of the schoolchildren was used as an integrative metric.

In addition, there was case study-based research that evaluated open-plan classrooms with respect to the learning and daily-life activities of schoolchildren. In a study on lesson development in open-space classrooms, Ueno found that certain configurations between the open classroom and the adjoining open space led to new kinds of lesson developments. Ueno then presented architectural recommendations with respect to furniture layouts that can support such lesson development, and the need for flexible arrangements.

Owing to the occurrence of novel lesson developments in this way, Kunigami et al and Kurakazu et al conducted studies to determine the process that follows. According to Kunigami et al, although the perceptions of open spaces were no less favorable in comparison with to those from when the open-plan design was introduced, the motivation and opportunities to use them declined, and the perceptions of classrooms that were opened up to the open space were unfavorable. Kurakazu et al conducted a fact-finding survey on the perception teachers had of open-plan designs. The results highlighted combinations of architectural modalities (the extent to which classrooms are open) and instructional strategies (learning groups) that are conducive to learning development. It was suggested that open-plan classrooms can be effectively applied when the number of schoolchildren in the grade is not excessive, and when each classroom has secured the freedom to close itself off depending on the circumstances. These studies provided useful guidelines for how lessons can be optimally tailored around an open-plan design. However, though this research explored the effects of open-plan classrooms on teaching styles, it did not consider the effectiveness of open-plan classrooms with respect to classroom management.

One study that evaluated open-plan classrooms with respect to classroom management was conducted by Hirata, who surveyed elementary schoolchildren in standard classrooms and those in open-plan classrooms and found that the latter were more motivated in their lesson activities and had more favorable attitudes toward class order. However, Hirata found certain opinions suggesting that there were no differences in academic performance or behavior, and refrained from drawing firm conclusions about the effects of open-plan classrooms. In this study, the effects of open-plan classes on schoolchildren are evaluated based on the extent to which schoolchildren are able to concentrate on their lessons. Rather than focusing on specific behaviors among the children, we asked teachers to give their general perceptions of the concentration level in their classrooms.

Methods of the study

The inconsistency in the evaluations of open-plan classrooms is the result of the complex interactions between different conditions. It also reflects the fact that research has poorly defined what is being evaluated. In a study on distraction and privacy in classrooms, Ahrentzen and Evans found that secluded areas were physically separated not only by wall enclosures but also by changes in floor levels. It was therefore stated that it is difficult to identify discrepancies in how classroom space is perceived based on environmental features. We attempted to categorize levels of openness based on how the classrooms are physically separated, not on the types of classroom facilities or structures. Based on the type of physical separation between the classroom and hallway, 4 categories of open-plan classrooms were formulated and analyzed in addition to a fifth category describing the standard (closed) type (see Figure ).

Some of the existing studies took the approach of asking teachers about the level of distraction among the children. In a similar manner, we referred to the perceptions teachers had of the concentration levels in their classes and the extent to which they found certain factors to have a distracting effect. In this way, we aimed to evaluate how differences in classroom designs affect the overall concentration level of the children.

Survey outline Survey purpose

In a questionnaire, we asked teachers to state their general perceptions of the levels of concentration in their classrooms to determine whether differences in classroom designs affect the concentration level of the children. Where such discrepancies were found to exist, the factors inhibiting the concentration level were examined.

Study area and participating schools

The area surveyed was Z City, near Tokyo. Moreover, Z City was one of the first municipalities in Japan to start building elementary schools with open-plan classrooms. In comparison with other areas, Z City has many open-plan elementary schools that are under the jurisdiction of the board of education. It was therefore decided that Z City would be ideal for quantitative analysis. A preliminary survey (eg, classroom observation) was conducted at 15 of the elementary schools in Z City, and the schools were asked to participate in the questionnaire. Of these schools, 14 agreed to take part; and the questionnaire was therefore administered to teachers in these 14 schools and in 2 traditional elementary schools resulting in a total of 16 elementary schools. The schools were generally within a 25 km diameter area (see Figure ). Some of the school districts overlapped with farmland and others were near a large commercial area, but all the schools were in residential areas.

Survey method and period

In the preliminary survey, we inspected the open-plan classrooms, conducted participant observation during lessons, and interviewed the school principals and other staff members about the physical school environment. We then developed a questionnaire, referring to the findings of the preliminary survey. We asked the principals of each school to distribute the questionnaires to the teachers. The teachers voluntarily and anonymously completed the questionnaires, and the schools were visited at a later date to collect the completed questionnaires . The questionnaires were distributed and collected during 2 different periods: the first was during February 2012 and March 2012 (8 schools), and the second was during February 2014 and March 2014 (8 schools). The main question items are presented in Table .

Main questionnaire items

Q1. Attributes

About the teachers Gender Length of the professional teaching experience Experiences of open-plan classrooms, etc. About the classes in charge Grade Number of the children in the class Number of the children in need of educational special supports, etc.

Q2. Usage conditions of the classrooms

Q3. Conditions of the children during lessons

General perceptions of the concentration degree Factors and frequency of distractions Frequency of distractions related to the children's actions, etc.

Q4. Facilities of teaching

Q5. Opinions about open-plan classrooms

Survey results Response

A total of 240 questionnaires were collected from the teachers, resulting in a response rate of 85.3%. Schools with return rates lower than 50% were excluded from the analysis. Accordingly, one school that returned 10 questionnaires was excluded. Moreover, one more questionnaire was excluded from a standard (closed)-type classroom, which was included in an open-plan school. The latter school was unique in that it had a semi-open type design but had a couple of exceptional standard-type classrooms. After excluding these 11 questionnaires, we were left with 229 returned questionnaires, a response rate (with respect to the remaining schools) of 87%. The total number of children in each of the remaining 15 schools ranged from 220 (8 classes) to 925 (26 classes), for an average of 591.6 children (17.7 classes).

Distribution of classroom designs and distribution of returned surveys by grade and teacher experience

Figure  presents the distribution of the classroom design types among the analyzed schools, and Table  presents the distribution of the collected questionnaires by the grades taught by the relevant teachers. Responses were received from schools that use all types of classroom designs . With respect to teaching experience (in years), the number of teachers with 10-20 years of teaching experience was somewhat low, but not sufficiently low to bias the sample.

Distribution of the questionnaire collection according to the classroom plan and the grade

[Table omitted. See PDF]

Analysis and discussion Class concentration level and the relevant variables

The variables that were assumed to affect the concentration of children in the classrooms (as perceived by the teachers) were investigated. This was achieved by analyzing their effects on the responses to question 3: “general perception of class concentration degree.” This item was rated using a 4-point scale: 1. the children concentrate very well; 2. the children concentrate fairly well; 3. the children tend to get distracted; and 4. the children cannot concentrate . The first 2 responses were defined as high concentration and the latter 2 as low concentration.

Number of children in class (class size) and concentration level

Among the effective responses, the number of children in the classrooms ranged from 19 to 40, with a mean of 32.6, a median value of 33.0, and a standard deviation of 4.02. The average class size in Z City was 27.5 children . Using this figure as a reference value, the results for class size were sorted into above-reference and below-reference categories, and the correlations with class concentration levels were analyzed (see Table ). No significant difference was observed in the distribution of the concentration degree; teachers in larger-sized classes were not less likely to give a “high concentration” response.

Degree of the concentration of children according to the number of children in a class

[Table omitted. See PDF]

P = 0.312, Pearson χ² = 1.024, df = 1.

Number of children with special needs and concentration level

Among the effective responses, the number of children with special needs in the classrooms ranged from 0 to 15, with a mean value of 3.0, a median of 3.0, and a standard deviation of 2.5. Using this median of 3.0 as a reference value, the results for the number of children with special needs were sorted into above-reference and below-reference categories, and the correlations with class concentration levels were analyzed (see Table ). It was found that teachers whose classes contained a relatively large number of children with special needs (4 or more) were significantly less likely to give a “high concentration” response.

Degree of the concentration of the children according to the number of children in need of special supports in a class

[Table omitted. See PDF]

P < 0.001, Pearson χ² = 18.685, df = 1.

⁴Significantly high.

However, an exact comparison was not possible, owing to discrepancies in the overall number of the children in the classrooms. The percentage of children in classrooms with special needs was therefore used in the analysis.

Ratio of children with special needs and concentration level

Among the effective responses, the percentage of children in the classrooms with special needs ranged from 0.0% to 42.9%, with a mean value of 9.4%, a median of 7.7%, and a standard deviation of 7.9%. Using this median of 7.7% as a reference value, the results of the proportion of children with special needs were sorted into above-reference and below-reference categories, and the correlations with class concentration levels were analyzed (see Table ). A significant difference was observed in the distribution of the concentration levels, whereby teachers whose classes contained a relatively high percentage of children with special needs were significantly less likely to give a “high concentration” response. This finding implies that the perception a teacher has of the concentration level of the class differs depending on the proportion of children in the class with special needs.

Degree of the concentration of children according to the percentage of children in need of special supports in a class

[Table omitted. See PDF]

P = 0.001, Pearson χ² = 10.946, df = 1.

⁶Significantly high.

Teaching experience and concentration level

The results were divided by the respective experience of the teachers (in years), and the correlations with class concentration levels were analyzed (see Table ). No significant difference was observed in the distribution of the concentration degree; years of teaching experience did not correlate with the likelihood of giving a “high concentration” response.

Degree of the concentration of children according to teaching experience

[Table omitted. See PDF]

P = 0.220, Pearson χ² = 4.420, df = 3.

Classroom design and concentration level

The results were by the types of classroom designs presented in Figure , and the correlations with class concentration levels were analyzed (see Table ). A significant difference was observed in the distribution of concentration levels whereby teachers of semi-open-type classes were significantly less likely, and teachers of removable-wall-type classes significantly more likely, to give a “high concentration” response. This finding suggests that classroom design is related to the perception the teacher has of the concentration level of their class .

Degree of the concentration of children according to the plan type of the classroom

[Table omitted. See PDF]

P = 0.001, Pearson χ² = 18.058, df = 4.

**P < 0.01, () = Adjusted residual.

Pairwise rankings were performed between pairs of classroom designs for which the distribution of the concentration degree significantly differed based on the Steel-Dwass test (see Table ). In comparison with the standard-type classroom, the semi-open-type classroom resulted in a significantly different distribution of the concentration degree. On the contrary, the concentration degree for the full-open-type classroom was not significantly different to that of the standard type.

Combinations of plans that display significant differences in the degree of the concentration of the children

Degree of concentration		Steel-Dwass P-value
High	Low
RW (removable-wall)	SO (semi-open)	0.010	<0.05
RW (removable-wall)	SP (sliding-panel)	0.015	<0.05
ST (standard)	SO (semi-open)	0.036	<0.05

Conclusions regarding class concentration level and variables

In Section , the variables that were assumed to affect the class concentration levels (as perceived by the teachers) were analyzed by comparing their effects on the perceptions the teachers had of their class concentration levels. The results indicated that the class concentration level is not affected by class size and years of teaching experience. However, it is affected by the number and percentage of the children in the class with special needs, and the classroom design. Teachers of classes with a high number or percentage of children with special needs were significantly less likely to perceive their children as concentrating well.

With respect to classroom design, the teachers of semi-open-type classes were significantly less likely, and teachers of removable-wall-type classes significantly more likely, to give a “high concentration” response. Moreover, when the perceived concentration levels among each classroom design type was compared, it was found that the perceived concentration levels of semi-open-type classes were significantly lower than those of standard-type classrooms. This finding suggests that the classroom plan type influences the perception of the class concentration level to a certain extent. However, it is assumed that this effect is not solely attributable to the physical degree of the openness of the classroom.

In the sample, there was no significant discrepancy either in the numbers or in the percentages of children with special needs among the classroom design types . Thus, the posit number/percentage of children with special needs and the classroom design type can be set as independent variables of the perception of the class concentration level. From an architectural perspective, it would be effective to evaluate the relationship between each classroom design type and class concentration level. Next, the relationship between the classroom design type and the extent to which teachers identify certain factors as distracting was examined.

Relationship between classroom design type and sources of distraction

Ahrentzen and Evans highlighted several classroom features that may distract students, including audio and visual factors, and factors related to the behavior of the children. These 3 categories were adopted for the examination of the sources of distraction in the classrooms. Several situations were identified, which teachers frequently found to be a source of distraction, and which were associated with a poor general perception of the class concentration level (see Table ). The correlations between these factors and classroom design types were analyzed. The responses were rated using a 4-point scale: 1. this hardly ever proves to be a distraction; 2. this rarely proves to be a distraction (several times a term); 3. this sometimes proves to be a distraction (around once a week); 4. this often proves to be a distraction. The first 2 responses were defined as not often and the last 2 as often.

Questionnaire items related to the concentration of the children

1) Related to the sound

• Frequency of distraction caused by external noise

2) Related to the sight

• Frequency of visual distraction caused by the people passing by the classroom

3) Related to the children's action

• Frequency of children entering the classroom after the class has started

• Frequency of the movement of the children to the surrounding area of the classroom without permission during lessons

• Frequency of children leaving their seats without permission during lessons

Distraction related to sound

The frequency with which external noise proves to be a distraction was compared according to the classroom design types (see Table ). It was found that external noise distracts the class significantly more often in full-open-type and semi-open-type classrooms, and that the frequency of distraction correlates with the physical openness of the classroom. This finding implies that, owing to the physical properties of sonic transmission, the more physically open a classroom is, the more frequently external noises distract the class. It is therefore reasonable to assume that efforts to improve the sonic environment at a physical level will have positive effects on the class concentration level.

Frequency of distraction caused by external noise

[Table omitted. See PDF]

P < 0.001, Pearson χ² =50.0318, df = 4.

*P < 0.05, **P < 0.01, () = Adjusted residual.

Pairwise rankings were performed between pairs of classroom designs for which the distribution of the concentration degree significantly differed based on the Steel-Dwass test (see Table ). In comparison with those who teach in standard-type classrooms, teachers who teach in classrooms that are regularly open (full-open type, semi-open type, and sliding-panel type) were significantly more likely to answer “often.” This finding indicates that opening up a classroom increases the frequency with which external noises distract the class, implying that it is important to find ways to improve the physical sonic environment.

Combinations of plans that display significant differences in the frequency of distraction caused by external noise

Frequency		Steel-Dwass P-value
Not often	Often
ST (standard)	FO (full-open)	0.000	<0.01
ST (standard)	SO (semi-open)	0.000	<0.01
ST (standard)	SP (sliding-panel)	0.000	<0.01
RW (removable-wall)	FO (full-open)	0.003	<0.01
RW (removable-wall)	SO (semi-open)	0.003	<0.01

Distraction related to sight

The frequency with which people passing by the classroom are a distraction was compared with respect to classroom design types (see Table ). It was found that passers-by distract the class significantly more often in full-open-type and sliding-panel-type classrooms, and significantly less often in standard-type and removable-wall-type classrooms.

Frequency of distraction caused by the people passing by the classroom

[Table omitted. See PDF]

P < 0.001, Pearson χ² = 37.000, df = 4.

*P < 0.05, **P < 0.01, () = Adjusted residual.

Pairwise rankings were performed between pairs of classroom designs for which the distribution of the concentration degree significantly differed based on the Steel-Dwass test (see Table ). As with the previous factor, when compared with those who teach in standard-type classrooms, teachers who teach in classrooms that are regularly open (full-open type, semi-open type, and sliding-panel type) were significantly more likely to answer “often.” Open-plan classrooms typically feature furniture, movable partitions, sliding-panels, and other contrivances designed to block part of the view of the children. This finding suggests that these features fail to significantly improve class concentration levels. Unlike the case with audible distractions, with respect to visual distractions, merely partitioning part of the classroom perimeter is not sufficient to minimize the distraction of visual stimuli. Thus, steps must be taken to ensure that the children never perceive the distracting visual stimuli in the first place. For example, children could be seated with their backs facing to the hallway.

Combinations of plans that display significant differences in the frequency of visual distraction caused by the people passing by the classroom

Frequency		Steel-Dwass P-value
Not often	Often
ST (standard)	SP (sliding-panel)	0.000	<0.01
ST (standard)	FO (full-open)	0.000	<0.01
RW (removable-wall)	SP (sliding-panel)	0.001	<0.01
ST (standard)	SO (semi-open)	0.001	<0.01
RW (removable-wall)	FO (full-open)	0.007	<0.01
RW (removable-wall)	SO (semi-open)	0.020	<0.05

The reason why sliding-panel-type classrooms were associated with a high distraction frequency is that when panels are moved to open up the classroom perimeter, there nothing blocking the view of the children. Under such a condition, it is difficult to prevent children from being distracted by visual stimuli, such as someone suddenly passing by. Even when the panels are situated so that the classroom perimeter is closed, distractions can still occur. For example, when panels are moved, gaps between panels sometimes emerge. Additionally, from the perspective of the children, there is uncertainty as to when nearby panels may suddenly slide away. As a result, children, particularly those sitting nearest the hallway, may tense up when they sense someone passing by. Thus, when designing sliding-panel-type classrooms, it may be effective to make the classroom of such a size or shape that children can be seated at a reasonable distance from the classroom perimeter.

Distractions related to the children's actions

Three sources of distraction related to the behavior of the children were examined, as follows. The teacher-perceived distraction frequencies of these sources were compared with the classroom design types.

Children entering the classroom after the class has started

The frequency with which children failing to enter the classroom by the start of the lesson proves to be a distraction was compared with respect to the classroom design type (see Table ). It was found that this behavior distracts the class significantly more often in semi-open-type classrooms, and significantly less often in removable-wall-type classrooms.

Frequency of children entering the classroom after the class has started

[Table omitted. See PDF]

P < 0.001, Pearson χ² = 22.407, df = 4.

*P < 0.05, **P < 0.01, () = Adjusted residual.

Using the Steel-Dwass test, pairwise rankings were performed between pairs of classroom designs for which the distribution of the concentration degree significantly differed (see Table ). Teachers who teach in semi-open-type classrooms were significantly more likely to answer “often,” when compared with those who teach in standard-type, removable-wall-type, and full-open-type classrooms.

Combinations of plans that display significant differences in the frequency of children entering the classroom after the class has started

Frequency		Steel-Dwass P-value
Not often	Often
RW (removable-wall)	SO (semi-open)	0.000	<0.01
ST (standard)	SO (semi-open)	0.000	<0.01
FO (full-open)	SO (semi-open)	0.006	<0.01

One reason why semi-open-type classrooms were associated with a high distraction frequency could be that it is difficult for children to get a clear sense of the boundary between the classroom and the hallway. Upon observing lessons, certain contrivances in full-open-type classrooms that were designed to compensate for the lack of any structural boundary were noted. For example, classroom furniture was laid out along the periphery, or the periphery was marked with tape, to clearly indicate the boundary to the children (see Figure , left photo) . On the contrary, semi-open-type classrooms are demarcated by moveable panels . In some cases, the panels may be laid out along the classroom periphery perpendicular to the axis of the periphery. In such cases, it can be difficult to determine whether the boundary lies at the near edge or far edge of a perpendicularly arranged panel. Additionally, in cases where the panels can be freely repositioned at a certain distance into the hallway, the boundary of each classroom to the adjoining hallway becomes inconsistent, making it difficult to determine the spatial extent of the hallway (see Figure , right photo). The high degree of the planning freedom in semi-open classrooms, whereby there is no exact point at which the classroom ends and the hallway begins, might make it more difficult for children to get a clear sense of the classroom space. If this is so, then it might explain why teachers of such classrooms are more likely to frequently experience cases wherein children are not in the classroom by the start of the lesson.

Children straying outside the classroom during a lesson

The frequency with which children straying outside the classroom during a lesson proves to be distracting was compared with respect to the classroom design type (see Table ). Given that 30% of the items exhibited an expected value of less than 5 in a chi-squared test, we presented only the distribution frequencies as a reference. Moreover, the Steel-Dwass test results are presented in Table .

Frequency of the movement of the children to the surrounding area of the classroom without permission during lessons

[Table omitted. See PDF]

Combinations of plans that display significant differences in the frequency of the movement of children to the surrounding area of the classroom without permission during lessons

Frequency		Steel-Dwass P-value
Not often	Often
ST (standard)	SO (semi-open)	0.000	<0.01
RW (removable-wall)	SO (semi-open)	0.015	<0.05
ST (standard)	FO (full-open)	0.021	<0.05
ST (standard)	SP (sliding-panel)	0.031	<0.05
FO (full-open)	SO (semi-open)	0.036	<0.05

According to the Steel-Dwass test, in comparison with those who teach in standard-type classrooms, teachers who teach in classrooms that are regularly open (full-open type, semi-open type, and sliding-panel type) were significantly more likely to answer that the children were distracted “often.” Semi-open classrooms, in particular, were associated with an “often” response, even in comparison with full-open-type classrooms. This finding indicates that though regularly open classrooms feature this source of distraction more often than standard-type classrooms, there is no significant correlation between the distraction frequency and the degree of classroom openness.

The reason why semi-open-type classrooms were associated with a high distraction frequency may be the lack of a clear sense of boundary between the classroom and the adjoining hallway, like the previous section (see Figure ). Another reason might be that arranging panels perpendicularly or at sporadic intervals may create areas that are invisible to the teacher but visible to the children. With such blind spots, the teacher may find it more difficult to monitor some of the children, which could increase the potential for children to stray outside the classroom. Another issue relates to the fact that the work of the children is displayed on the panels. Thus, a child who pauses to look at the work on the panels, even if only momentarily, may end up gazing out across the hallway, naturally creating the temptation to stray outside the classroom (see Figure ).

Children roaming around the classroom during a lesson

The frequency with which children roaming around the classroom during a lesson proves to be distracting was compared with respect to the classroom design type (see Table ). It was found that this behavior distracts the class significantly more often in semi-open-type classrooms. Moreover, no significant differences were observed with other classroom design types.

Frequency of children leaving their seats without permission during lesson

[Table omitted. See PDF]

P = 0.019, Pearson χ² = 11.811, df = 4.

*P < 0.05, () = Adjusted residual.

According to the Steel-Dwass test, teachers of semi-open-type classrooms were significantly more likely to answer that their students were distracted “often,” when compared with to those who teach in standard-type and removable-wall-type classrooms, and those who teach in full-open-type classrooms (Table ).

Combinations of plans that display significant differences in the frequency of children leaving their seats without permission during lesson

Frequency		Steel-Dwass P-value
Not often	Often
ST (standard)	SO (semi-open)	0.001	<0.01
FO (full-open)	SO (semi-open)	0.017	<0.05
RW (removable-wall)	SO (semi-open)	0.014	<0.05

The reason why semi-open-type classrooms were associated with a high distraction frequency, may be related to the dense seating arrangements. When semi-open-type classrooms use a moveable panel, the boundary with the adjoining hallway becomes partly partitioned and partly open . In such classrooms, the seating tends to be clustered around the section of the classroom that is level with the partitioned section (see Figure ). Consequently, despite the classroom size not being significantly different from that of standard-type classrooms, semi-open classrooms tend to have a markedly denser lesson-time seating arrangement. A dense seating arrangement may be a psychological factor, that is, a factor that makes children want to leave their seats during a lesson. This seating arrangement appears to be typical to semi-open-type classrooms; as we did not observe the trend among full-open-type classrooms, which are entirely unpartitioned, or among sliding-panel-type classrooms, in which all the periphery sections can be opened or closed.

Conclusions regarding classroom design type and the sources of distraction

One of the factors that was assumed as a possible distraction to the class was external noise. It was found that the reported frequency with which the class gets distracted by external noise correlates with the physical openness of the classroom. Thus, methods insulate open-plan classrooms from external noise are required.

With respect to visual distraction, the analysis of the reported frequency with which children get distracted by someone passing by the classroom suggested that masking parts of the classroom that are open to the hallway is an ineffective strategy. A more effective strategy is to find ways to ensure that the children never perceive the visual stimuli in the first place, such as by having the children face away from the hallway. The analysis also highlighted that visual distractions occur frequently in sliding-panel-type classrooms, which can be entirely partitioned from the outside. It was concluded that children in these classrooms may get additionally distracted as a result of the psychological tension that is typical to this design type (the panel next to you may be slid open at any moment), and in cases wherein the panels are left in an open position such that there is nothing that blocks out visual stimuli.

The reported frequency with which the children get distracted by behavioral distractions was also analyzed.

Upon analyzing the reported frequency with which a class gets distracted as a result of children not being in class by the start of the lesson, it was found that this behavior is more problematic in semi-open-type classrooms than in full-open-type classrooms. From this finding, it was concluded that this distraction factor is related not only to the openness of the classroom but also to children not having a clear idea of where the classroom ends and the hallway begins.

Another behavioral distraction was children straying outside the classroom during a lesson. The lack of a partition may increase the frequency of such a distraction, as is indicated by the fact that this behavior occurs more frequently in classrooms that are regularly open (full-open type, semi-open type, and sliding-panel type). Given that this behavior is even more frequent in semi-open classrooms than in full-open-type classrooms, it may in many cases be related to the typical characteristics of semi-open classrooms, including ambiguous classroom-hallway boundaries and the potential for blind spots (from the perspective of the teacher) along the periphery. Moreover, an issue related to the flexibility in the layout of moveable panels was observed, namely, that panels may be positioned in such a way that children who gaze at works on them are naturally tempted to stray from the classroom.

As for children roaming around the classroom during lessons, this behavior occurred significantly more frequently in semi-open-type classrooms, even when compared with full-open-type classrooms. One particular feature of semi-open-type classrooms is that only a part of the classroom-hallway boundary is partitioned, and seating arrangements tend to be clustered in the part of the room level with this partitioned section. Hence, it was observed that children in these classrooms were seated much closer together during lessons, when compared with to those in other classroom types. It was surmised that this dense seating arrangement is a psychological factor, as children are less comfortable being seated so close together, and are therefore more inclined to leave their seats.

Conclusion

In this study, 229 teachers from 15 public elementary schools in Z City were surveyed. The effects of the differences in classroom design on the concentration level of the children as perceived by the teacher (ie, whether they influence the responses for the questionnaire item “general perception of class concentration level”) was examined. The factors that might explain the differences in the results were then examined.

The perceptions the teachers had of the class concentration level were unaffected by some of the teacher/class attributes, such as the teaching experience (in years) and the number of children in their class, but they were influenced by the number and proportion of children in their class with special needs. They were also influenced by the classroom design type. Gump noted that open-plan classrooms can adversely affect the concentration levels of children. However, no significant difference was observed in the teacher-perceived concentration levels between full-open-type classrooms and standard-type classrooms. Moreover, children in semi-open-type classrooms had significantly lower concentration levels. This finding suggests that the perceptions teachers had of their class concentration levels do not correlate with the degree to which a classroom is open. Moreover, class concentration is influenced by the psychological and behavioral properties associated with the particular classroom-hallway border structure of the classroom design in question. For example, the removable-wall-type classroom, which is regularly partitioned from the hallway, demonstrated similar results to the standard-type classroom, when compared with other open-plan classroom types (see Tables , , , , and ). The sliding-panel-type classroom, which can also be partitioned, exhibited a different trend.

The psychological and behavioral aspects of the classroom and possible factors that inhibit concentration were then considered. Specifically, the teacher-reported frequency with which each factor occurs by classroom design type. With the exclusion of noise-related distraction, none of the distraction factors correlated (in terms of how often they occur according to the teachers) with the openness of the classroom periphery. With respect to external noise, the more open a classroom was, the more frequently this was seen to distract the class, revealing the necessity for physical improvements, to insulate open-plan classrooms from external noise. With respect to visual factors, physically masking parts of the classroom is an ineffective strategy, as suggested by the results. A more effective approach to blocking out distracting visual stimuli is to design classrooms with a shape and space that facilitates certain seating arrangements, such children seated with their backs to the classroom-hallway boundary, or at certain distances from the boundary. Moreover, with respect to visual distraction, the frequency of behavioral factors was not influenced only by the physical openness of the classroom. Other elements contributed to these behavioral factors, including children not having a clear sense of the classroom-hallway boundary, non-uniform partitioning, and a seating arrangement trend related to such factors. While the findings should be examined carefully in a separate systematic study, they constitute referential data for those involved in improving the current environments of school classrooms.

Acknowledgments

In undertaking this study, we were supported to a great extent by individuals from the board of education and the relevant schools of Z City. We express our gratitude to these individuals.

Disclosure

The authors have no conflict of interest.

Notes:

¹⁰⁰³Rather than simply measuring the degree of the open perimeter, we referred to the degree of openness when viewed horizontally. Specifically, we determined the category based on the number of view-obstructing features (such as a wall or furniture) between the floor and the ceiling. Certain classrooms had 90 cm wing walls or studs; and given that these were small in size, we did not consider them as barrier objects.

¹⁰⁰⁴We asked the teachers to seal their completed questionnaire forms in an envelope and place them in a collection box.

¹⁰⁰⁵Only one of the returned forms was from a third grade teacher at a school with a removable-wall-type design; however, we did not view this as problematic, considering that we received 7 such forms from fourth grade, which is in the middle grade same as third grade, teachers.

¹⁰⁰⁶We considered that the behavioral indicators teachers refer to in judging how well the class is concentrating may differ depending on the lesson content, class composition, and teaching plan. Accordingly, we did not present any classroom behavior indices for the teachers to refer to. Instead, we asked the teachers for their subjective judgments. We used a larger sample size and conducted statistical processing to absorb any individual differences in judgment.

¹⁰⁰⁷The national average, which includes underpopulated areas, is 22.4. Thus, class sizes in Z City, an urbanized area, are relatively high.

¹⁰⁰⁸Of the teachers of open-plan classrooms, there were 122 for whom we could surmise both their teaching experience (in years) in an open-plan classroom, and in a standard-type classroom. Of these, 80 reported that they had previously taught in a standard-type classroom, whereas 42 reported that they had not. When we compared these 2 sets of teachers, we found no significant difference between them in terms of their perceptions of the concentration levels. Of the 32 teachers who were teaching in a standard-type classroom, 4 had taught in an open-plan classroom in the past and 28 had not. Given the small number of teachers, we were unable to compare these 2 sets of teachers in terms of their perceptions of concentration level.

¹⁰⁰⁹We used pairwise rankings, derived by the Steel-Dwass test, to perform multiple comparisons between pairs of items. The criterion variable (class concentration level in 4.1, distraction frequency in 4.2) was rated using a 4-point scale.

¹⁰¹⁰Using the Dunnett test, we compared standard-type classrooms and semi-open-type classrooms with other types of classrooms, with respect to the average number and percentage of children with special needs. The results did not reveal any significant differences in the averages (number of children with special needs in a class: control=semi-open: P = 0.58-1.00, control=standard: P = 0.71-1.00; percentage of children with special needs in a class: control=semi-open: P = 0.75-1.00, control=standard: P = 0.85-1.00).

¹⁰¹¹Both variables, the distraction frequency and the general perception of the class concentration level, were rated using a 4-point scale. Accordingly, we analyzed the results using Spearman's rho. The analysis indicated that the distraction frequency of the items in Table is associated with a poor perception of the concentration level (ρ = −0.28 to −0.50; P < 0.01 in each case).

¹⁰¹²In one of the schools we observed, the flooring materials of the classrooms differed from those of the adjoining hallway. In 3 of the schools, classroom furniture was laid out along the periphery. In the 2 other schools, 70% of the classrooms had their periphery with the adjoining hallway indicated by tape, or classroom furniture such as satchel lockers.

¹⁰¹³Of the 2 schools with semi-open-type classrooms, 1 used removable panels all year round. As for the other school, on the day-long surveys in April, July, November, and February, all the classerooms were using at least 1 panel.

¹⁰¹⁴The average class size among schools with standard-type classrooms was 31.6, and 30.5 among schools with semi-open-type classrooms.