Studies in Educational Evaluation 58 (2018) 167–177

Contents lists available at ScienceDirect

Studies in Educational Evaluation

journal homepage: www.elsevier.com/locate/stueduc

Comparative analysis of the impact of traditional versus innovative learning T

environment on student attitudes and learning outcomes
⁎
Terry Byers , Wes Imms, Elizabeth Hartnell-Young
The Anglican Church Grammar School and The University of Melbourne, The University of Melbourne, Australia

A R T I C LE I N FO A B S T R A C T

Keywords: This study explored the claim that different classroom layouts can affect teaching and learning. At the time of the
Classroom space study, there were few robust evaluative frameworks able to isolate and then measure the impact of different
Engagement educational layouts. In response, this study employed a quasi-experimental approached facilitated by a Single
Learning experiences Subject Research Design (SSRD) to compare two different classroom layouts - a traditional classroom layout and
Learning outcomes
‘Innovative Learning Environment’ (ILE) in an Australian secondary schooling context. The study compared
Quasi-experimental
Secondary school
students’ attitudes to their learning experiences, motivation, engagement and academic outcomes in each layout
over a school year. Comparative analyses highlighted how students’ attitudes to their learning experiences and
engagement differed in the two designs. A correlation was identified between enhanced student attitudes in an
ILE and higher English, Humanities and Mathematics academic achievement when compared with cognitively
matched peers who occupied a traditional classroom for the same period. This initial empirical evidence, even
though restricted to a single site, was able to discern a measurable link between the occupation of different
learning spaces and an impact on student learning experiences, engagement and academic outcomes in sec-
ondary schooling context.

1. Introduction
The form and function of what constitutes an effective learning
environment to meet emerging educational imperatives in many
countries have become a highly relevant and topical issue (Benade,
2016; Dovey & Fisher, 2014). There appears to be a rejection of existing
conventional or traditional classrooms in favour of more dynamic and
responsive Innovative Learning Environments (ILEs). The OECD (2013)
defines ILEs as multi-modal, technology-infused and flexible learning
spaces more responsive to the needs of twenty-first-century learners
than a traditional classroom. ILEs can range from adaptive, purposeful
spaces through to open-plan environments (Benade, 2016; Tanner,
2008). It is suggested that the affordances of ILEs will better support
those pedagogical changes that will facilitate learning experiences that
enhance students’ engagement and motivation (Ryan & Patrick, 2001)
and improve academic achievement (Tanner, 2009). As a consequence,
the focus on creating new school learning environments has become a
matter of strategic policy in Australia (Mulcahy & Morrison, 2017). The
systemic public funding from the Federal Government’s Building the
Education Revolution (BER) saw AUS $16.2 billion invested in delivering
24,382 projects in 9526 schools from 2009 to 2011 (Wall, 2009).
Despite the current interest and investment in school building
programs, authors lament the lack evidence outlining exactly how
different spatial layouts support, or hinder, the desired evolution of
teaching and learning (Brooks, 2011; Chandler, 2009; Gislason, 2010;
Higgins, Hall, Wall, Woolner, & McCaughey, 2005; McGregor, 2004).
Reviews of the literature around ILEs consistently cite few evaluative
frameworks and methods (Painter et al., 2013; Tanner, 2008) able to
elicit evidence concerning their pedagogical impact (Blackmore,
Bateman, O’Mara, & Loughlin, 2011; Gislason, 2010). Conversations
around the potential of these ‘new’ spaces often rely upon theoretical
expositions, case studies, or post-occupancy evaluations of tertiary
spaces. In the secondary schooling context, rigorous empirical evalua-
tion in how traditional classrooms or ILEs influence student learning
experiences, engagement, motivation and academic achievement is
lacking.
The impact of the different classroom layouts – a traditional class-
room and a retrofitted ILE – on teaching and learning were evaluated
through a year-long study in a secondary schooling context. A quasi-
experimental approach facilitated by a Single Subject Research Design
(SSRD) investigated the impact of these two layouts on students’ atti-
tudes to their learning (learning experiences, motivation, and engage-
ment). The student sample was allocated to three groups (ILE
Intervention, ILE Control, and Traditional Control). Between- (ILE

⁎
Corresponding author at: Anglican Church Grammar School, Oaklands Parade, East Brisbane, QLD, Australia.
E-mail address: Terry.Byers@churchie.com.au (T. Byers).

https://doi.org/10.1016/j.stueduc.2018.07.003
Received 20 October 2017; Received in revised form 5 April 2018; Accepted 11 July 2018
0191-491X/ © 2018 Elsevier Ltd. All rights reserved.
T. Byers et al.
Control and Traditional Controls) and within-group (ILE Intervention)
comparative analysis investigated if aspects of learning differed due to
the different classroom layout (traditional and ILE) and not some other
confounding effects such as assessment, curriculum, class composition,
cognitive ability and the teacher. Thereby, any variations in academic
achievement could be better correlated to these changes to the class-
room layout.
Comparative analysis of student survey and assessment data in
English, Humanities, and Mathematics subjects indicated how student
attitudes and academic achievement differed in the two layouts. Both
between- and within-group analysis indicated that those students who
occupied an ILE reported that they experienced a greater incidence of
student-centric modalities of learning (i.e., active learning; collabora-
tive learning; independent creativity and personalisation) than their
peers experienced in a traditional layout. The relationship was not
causal. Statistical analysis highlighted the mediating influence of par-
ticular teachers who were more able to exploit the additional affor-
dances of the ILE for pedagogical gain. For these classes, changes to the
nature of learning experiences were related to a higher assessment of
student cognitive and emotional engagement in learning. Together it is
suggested that the changes experienced by students in an ILE compared
to their cognitively matched peers in a traditional layout, correlated to
improved academic achievement. The findings and general conclusions
presented here replicate the evidence presented by earlier spatial in-
terventions at this site (see Byers, 2017; Byers and Imms, 2014; Byers
et al., 2014; Byers and Imms, 2016; Imms and Byers, 2016; Imms et al.,
2017). A replication across studies using a similar design and methods
within the same context, but different student samples, builds a case
around the pedagogical impact of different school learning spaces.
2. Literature review
2.1. The deficiency of the traditional classroom layout
The conventional or traditional classroom model, which is still oc-
cupied by the vast majority of students on a daily basis, emerged from
the modernist architectural period after the Second World War (Dovey
& Fisher, 2014). A recent study by Byers et al. (2018) found a corre-
lation with a higher incidence of those teacher-centric pedagogies (i.e.
teacher facilitated direct instruction or small group discussion/in-
struction) in schools where traditional classroom layouts were the
dominant archetype. This correlation could be due to the influence of
built pedagogy. Reynard (2009) suggested that the built pedagogy of
the traditional classroom's layout, established through its fixed in-
structional setting facing the front teaching ‘fireplace’ position, makes it
natural for the teacher to stand and deliver content transmitted through
didactic instruction. Furthermore, as Dovey and Fisher (2014) argued,
the rigidity of its layout inhibits the ability of teachers to enact a wider
spectrum of pedagogies that are more likely to facilitate student-centric
and technology-enhanced learning as favoured in current educational
policies. Proponents of ILEs suggest, but with little tangible evidence,
that traditional classrooms are more likely to support those pedagogies
associated with surface learning experiences and not those processes
required for deep learning (Dumont & Istance, 2010; Hattie &
Donoghue, 2016).
2.2. How innovative learning environments affect student learning
experiences and engagement
In countries such as Australia, New Zealand and the United States of
America, the significant investment in the infrastructure of ILEs is
perceived as a necessary, or optimal means, to facilitate the twenty-
first-century learning and skills (Dovey & Fisher, 2014; Mulcahy &
Morrison, 2017). Typically, this view of learning is associated with an
approach to learning that is more multiplicitous (or a wide variety),
technology-mediated, and student-centric thought to best equip and
168
Studies in Educational Evaluation 58 (2018) 167–177
prepare students for their future working life (Dovey & Fisher, 2014;
Mulcahy, 2016). Mulcahy (2016, p. 19) describes how the growing
“official policy discourse of twenty-first-century learning, twenty-first-
century capabilities and personalised learning” has become intertwined
with the current focus on the physical learning environment.
Underpinning the narrative around ILEs is that they are somehow
better in supporting the shift from teacher-centric to more student-
centred learning (Dumont & Istance, 2010; Mulcahy, 2016). It is sug-
gested that the built pedagogy of the ILEs removes those inherent
spatial barriers, such as the teaching fireplace and rigid seating layout,
supporting teachers to shift away from overtly didactic pedagogies
(Reynard, 2009). Rather than curtail learning to a single pedagogical
mode, the affordances typically associated with an ILE are thought to
respond to and accommodate a much wider range of learning mod-
alities (Dovey & Fisher, 2014). Dumont and Istance (2010) suggest that
an ILE can support those modalities that support a more progressive
view of learners as active, collaborative, and constructive in their ac-
tivities. Ryan and Patrick (2001) proposed that the environments cre-
ated by these modalities of learning are correlated to improvement in
students’ engagement and motivation in learning.
2.3. Research designs and methods to evaluate school spaces
Various reviews have found few evaluative approaches available to
discern exactly how different school learning environments affect sec-
ondary student attitudes to their learning and impact on academic
achievement (see Blackmore et al., 2011; Cleveland & Fisher, 2014;
Painter et al., 2013; Tanner, 2008). Cleveland and Fisher (2014) note
that evaluative methods in the secondary schooling context often rely
on student and teacher perceptions to measure the social or psychoso-
cial effect of learning environments (see Aldridge, Fraser, Bell, &
Dorman, 2012; Dorman & Fraser, 2009; Fraser, 1982; Zandvliet &
Fraser, 2004). Here the concept of the learning environment refers to
the ambience, atmosphere, climate, or tone of a setting and its impact
on human behaviour (Dorman, Aldridge, & Fraser, 2006) and on the
instructional processes (Aldridge et al., 2012). Cleveland and Fisher
note that these studies use large-scale attitudinal questionnaires and
surveys, often comparing different research sites but rarely do they
focus on the spatial impact on academic achievement.
The review by Cleveland and Fisher (2014) identified other ap-
proaches and tools outside the domain of the social or psychosocial
evaluation of different school learning environments. The School
Building Rating Scale (Sanoff, 2001) and the Design Quality Indicators for
Schools (DQIfS) (CABE, 2005) are examples of tools that evaluate the
perceptions of the physical features of space itself (Cleveland & Fisher,
2014). They also identified the relatively new domain of Indoor En-
vironmental Quality (IEQ). The IEQ process ascertains the ‘environ-
mental performance’ of space by measuring the desirable physical
conditions (i.e., air quality, light, noise, spatial density, temperature,
and ventilation) that best support teaching and learning (Barrett,
Zhang, Moffat, & Kobbacy, 2013; Soccio, 2014). However, the focus of
these measures is on the physical features and performance of the
spaces and not how they influence those pedagogies and the learning
experiences thought to improve student academic achievement, en-
gagement, and motivation.
2.4. Earlier studies making a case for a change in classroom space
Reviews by Blackmore et al. (2011) and Painter et al. (2013)
identified few studies that make an empirically grounded case for
changes in classroom space. A small number of empirical studies have
sought to isolate the variable of the learning environment on academic
outcomes (see Barrett et al., 2013; Brooks, 2011; Tanner, 2008, 2009).
Even though in different contexts, their collective findings suggest that
the quality and nature of the classroom environment in various schools
correlated with differences in student academic achievement. However,
T. Byers et al.
their designs were unable to delineate what occurred in the different
spaces and how this influenced student attitudes to their learning and
academic outcomes.
To better understand this potential relationship and how it affected
student attitudes to their learning and academic achievement, previous
studies at the current site assessed the educative value of the physical
learning environment. These earlier studies (see Byers, 2017; Byers and
Imms, 2014; Byers et al., 2014; Byers and Imms, 2016; Imms and Byers,
2016; Imms et al., 2017) devised, tested, and refined a quasi-experi-
mental approach through an SSRD in a secondary school setting. The
approach accounted for some of the intervening variables inherent in
the school setting (such as assessment, curriculum, class composition
cognitive ability and teacher) to isolate the effect of a spatial inter-
vention (moving from a traditional classroom layout to an ILE).
Findings linked the occupation of an ILE with statistically significant
improvements in student attitudes in the effective utilisation of tech-
nology, the incidence of more active and responsive learning experi-
ences, and enhanced behavioural and cognitive engagement. These
changes correlated with statistically significant improvements in
English and Mathematics academic achievement (average g = +0.53).
Similar to the findings of Tanner (2008), Hierarchical Linear Modelling
calculated an averaged 9% variance in achievement attributed to the
different learning environments (when the confounding variables of
student cognitive ability, class composition, and the teacher were
controlled) (Byers, 2016). However, these studies were limited to a
within-group comparison of intervention groups. Therefore, they were
unable to mitigate various threats to the validity of conclusions due to
the ‘novelty of change.' As a result, the study that is the subject of this
article incorporates a longitudinal between-group comparison of con-
trol groups with a sample of considerably greater statistical power.
3. The study
This study evaluated the impact of the different classroom layouts –
a traditional classroom and a retrofitted ILE – on teaching and learning.
Over the period of a school year, between- and within-group com-
parative analysis of control and intervention groups evaluated if (and
how) student attitudes to their learning and academic outcomes were
affected by the different classroom layout (traditional and ILE) (see
Fig. 3 for study flowchart). Hence, what was of interest to this study
was:
1 To what extent do different classroom layouts affect student atti-
tudes to their learning experiences?
2 To what degree does the inhabitation of different classroom layouts
affect student attitudes to their motivation?
3 To what extent does the inhabitation of different classroom layouts
affect student attitudes to their engagement?
4 To what degree do different layouts affect student learning out-
comes, taking into account their cognitive ability and curriculum
material and assessment instruments?
3.1. The spaces
The study took place at a secondary school in a metropolitan city in
Australia, in three buildings constructed between the 1940s and 1960s.
The building contained 22 general learning area classrooms and ac-
commodated all year 7, 8, and 9 classes (11- to 15-year-olds) at the
school. Each space had a data projector and screen with the adequate
wireless infrastructure to connect student and teacher digital devices.
The traditional classrooms (Fig. 1) reflected a conventional design;
cellular spaces with desks and chairs arranged in fixed rows or small
group settings. The desks faced the fireplace teaching position at the
front of the room, delineated by a teacher desk, whiteboard, and data
projection. Reynard (2009) suggested that such a spatial arrangement
about the front fireplace emphasises a teacher-centric didactic
169
Studies in Educational Evaluation 58 (2018) 167–177
instructional approach.
The second type of space was a classroom that was retrofitted to
create an ILE (Fig. 2). In keeping with the philosophy behind ILEs, these
spaces utilised a combination of spatial (existing and new furniture and
additional whiteboards) and digital technologies (multiple teacher in-
puts into the data projector) to support teachers to influence and mould
the space to their pedagogical intent (Monahan, 2002). The key to the
layout was the de-emphasis of the front fireplace through creating a
polycentric (many focal points) layout. The polycentric layout altered
the built pedagogy of the space to encourage changes in behaviours that
would affect pedagogies and learning experiences. The cost of the ret-
rofit of the six classrooms ranged from $3950 to $4800 per space. The
nature of the retrofit presents a cost-efficient and transferrable solution
that could be easily transferred into any secondary, or even primary,
schooling context.
3.2. Sampling
The sample included students (n = 385) and teachers (n = 21) and
encompassed students from 22 classes from year 7 (n = 92), year 8 (n
= 158) and year 9 (n = 135). The overall student participation rate
(with parental consent) of the study was 64%. A priori power analysis
(0.8) with a probability level of (p = 0.05) and estimated Cohen’s ef-
fect size (d = 0.5) indicated that both the overall and year-level cohort
samples were adequate. Furthermore, the overall retention rate for the
study (82.4%) was above that specified in the quality indicators by
Gersten et al. (2005).
These classes were divided into three experimental groups based on
the existing timetable allocation (Table 1). The “ILE Intervention” (II)
group consisted of six classes that spent a semester in each classroom
layout. These classes were the focus of a within-group analysis to
measure the effect of the spatial change from a traditional (baseline) to
an ILE (intervention) space. The “ILE Control” (IC) group consisted of
six classes and spent the entire study (school year) in an ILE. The re-
maining ten Traditional Control (TC) classes stayed in an unchanged
traditional classroom for the same period. Normed and standardised
data from the Academic Assessment Services (AAS) cognitive (non-
verbal and verbal) ability instrument matched the control groups (IC
and TC) for between-group analysis. This large-scale psychometric in-
strument, normed to the respective Australian year-level student po-
pulation, acted as a proxy for student cognitive ability. Matched “like”
ability (high- and mixed-ability) were in fact statistically similar classes,
as determined by independent-group t-tests (p > 0.05).
3.3. Methods
The data collection methods consisted of an anonymous Linking
Pedagogy, Technology, and Space (LPTS) student attitudinal survey
instrument and student academic achievement and cognitive ability
data. The LPTS survey focused on answering research questions 1 to 3.
It consisted of 21 items assigned to five underlying scales. Table 2
shows scale descriptions and independent variables for each scale. The
scales evaluated the suggestion that the shift from traditional class-
rooms to ILEs affects discernable changes to teaching that affect a
greater incidence of student-centric learning that positively influence
learner motivation and engagement (Dumont & Istance, 2010).
Scales A and B examined if different spatial layouts (traditional and
ILE) affected student perceptions of the incidence and effectiveness of
those experience often associated with student-centric learning. At the
time, the literature suggested that ILEs are more attuned to supporting
learning that is more active, collaborative and creative than more tra-
ditional layouts. The survey replicated applicable elements of the
Tamim, Lowerison, Schmid, Bernard, and Abrami, (2011) study derived
from the APA Learner-Centred principles developed by Lambert and
McCombs (1988). Even though the Tamim et al. (2011) study focused
on evaluating technology-enabled learning environments, its
T. Byers et al. Studies in Educational Evaluation 58 (2018) 167–177

Fig. 1. Traditional Classroom Layout.

Fig. 2. Retrofitted classroom in the design of an Innovative Learning Environment (ILE).

assessment of the core elements of active, student-centred experiences
drew parallels with the claims made by proponents of ILEs.
Scales C, D and E evaluated the claim that changes to student-cen-
tric learning (Dumont & Istance, 2010) would correlate to improved
student engagement and motivation in learning (Ryan & Patrick, 2001).
Questions were derived from the ‘Motivated Strategies for Learning
Questionnaire’ (MSLQ). The MSLQ presented a proven and reliable
unidimensional student self-report of changes to motivational ex-
pectancies and values and engagement through instructional-focused
interventions (Fredricks, Blumenfeld, & Paris, 2004; Pintrich & De
Groot, 1990). Scale C evaluated motivational values by the examination
of student self-assessment of their intrinsic and extrinsic goal orienta-
tions and task value. Scale D assessed student control of learning beliefs
and self-efficacy for learning and performance to ascertain their
motivational expectancies. Finally, Scale E adapted questions that fo-
cused on cognitive and emotional engagement in learning in a current
learning environment.
Before analysis, LPTS survey data underwent reliability, normality
and sphericity tests. Measures of internal consistency, calculated
through Cronbach’s alpha, were within the acceptable range from 0.75
to 0.90 (Gliem & Gliem, 2003), The calculated alphas supported data
for each class, in each domain, to be summed and treated as one unit for
analysis (Table 3). With the sample size of less than 50, due to the class-
based unit of analysis, the application of Shapiro-Wilk’s statistic sup-
ported the assumption of normality for all scales (Ghasemi & Zahediasl,
2012). Finally, the application of Mauchly’s test for sphericity validated
the RM-ANOVA process by the examination of the equal variance of the
differences between all combinations of within-subject pairs. Instances,

170
T. Byers et al. Studies in Educational Evaluation 58 (2018) 167–177

Fig. 3. Study flow-chart of research phases and methods.

when sphericity could not be assumed, Greenhouse-Geisser-corrected
values were reported.
The final research question focused on evaluating whether different
learning environments affected student learning outcomes. Student
academic results, grades (on a 15-point scale from A+ to E−) awarded
by a criterion-based system for assignments and testsacted as a proxy
for learning outcomes. They represented a fair and unbiased dataset
with all students in each year-level and subject completing the same
assessment instruments of the same curriculum conducted under the
same conditions. Teachers marked these assessments according to
agreed marking schemas, with final judgements on student achieve-
ment decided through a collective moderation process.
3.4. Data analysis
RM-ANOVAs and post-hoc visual analysis evaluated student re-
sponses to the LPTS survey in two distinct phases. The first focused on
the repeated measures within-group analysis of the six intervention (II)
classes to assess the impact of the spatial change from traditional to ILE.
Baguley (2012) highlighted the tendency for positively correlated data
and sphericity to affect the findings from the within-group analysis of
repeated measures data. As a consequence, and using the
recommendation of Baguley (2012), the application of the Cousineau-
Morey process normalised the repeated measures data to correct in-
stances of both positively correlated data and sphericity.
The second phase facilitated the between-group analysis of cogni-
tively matched control (IC and TC) classes, through one between- and
one within-subjects factor. The between-group RM-ANOVA evaluated
the main effects on student attitudes to their learning (learning space
and time), The RM-ANOVA also focused on the possible interactions
(learning space x time) throughout the study to determine the potential
longer-term interaction of these effects. The analysis examined the
longitudinal influence of different classroom layouts by comparing the
responses of ‘similar’ (cognitively matched through the AAS ability
data) students over the school year, without the novelty brought by the
spatial intervention.
The addition of post-hoc visual analysis to RM-ANOVAs was
keeping with the SSRD. The visual analysis presented a known means to
facilitate the holistic evaluation of the behaviour of each class. The
process provided a detailed picture of each classes response, which
provided a complete understanding of specific idiosyncrasies and trends
present within and between classes. The visual analysis process readily
identified the possible influence of confounding variables (i.e. changes
in assessment or curriculum) on responses that were outside the scope

Table 1
Summary of the Sample Size, Participation, and Retention Rates.
Study Group Sample Size Participation Rate Retention Rate

Yr. 7 Yr. 8 Yr. 9 Yr. 7 Yr. 8 Yr. 9 Yr. 7 Yr. 8 Yr. 9

ILE Intervention (II) 37 21 58 69% 74% 68% 84% 74% 84%

ILE Control (IC) 28 92 NAa 52% 79% NA 86% 86% NA
Traditional Control (TC) 27 45 77 56% 56% 54% 81% 78% 89%

a
Not applicable – As the random allocation of classes meant no year 9 class was timetabled in an IC classroom at the start of the school year.

171
T. Byers et al. Studies in Educational Evaluation 58 (2018) 167–177

Table 2
Descriptive Information for the Linking Pedagogy, Technology and Space (LPTS) Survey.
LPTS Scale Independent Variables Sample Item

Learning experiences (Scale A)
Spatial Effect on Learning Experiences (Scale B)
Motivation Values (Scale C)
Motivation Expectancy (Scale D)
Spatial effect of Student Engagement (Scale E)
Teacher ownership My Teacher/s encourage collaborative learning and group work (2 to 4 students)
Personalisation
Active learning
Collaboration
Fostering creativity
Positive effect on teacher practice This space has had a positive effect on the way my teacher delivers lessons
Enabling Student-centered learning
Encouraging collaboration
Extrinsic goal orientation If I can, I want to get better results in this class than most of the other students
Intrinsic goal orientation
Task value
Control of learning beliefs Taking into account the difficulty of my subjects, the teachers and my skills, I think I will
Self-efficacy do well
Cognitive engagement This space makes me more interested in my learning
Emotional engagement

of this study. 4. Results and discussion

The criterion, informed by the work of Baguley (2012), for post-hoc
visual analysis consisted of; mean level change and trend; the im-
mediacy of the effect; overlap, trend, and variability of CIs. Fig. 4 de-
monstrates how the post-hoc analysis discerned if the statistically sig-
nificant changes determined by the RM-ANOVA calculations were due
to the spatial intervention (Class 7.1 analysis of Scale B on the left) and
not influenced by some other confounding variable (Class 7.2 analysis
of Scale B on the right).
A combination of within-group effect size calculations and
Hierarchical Linear Modelling (HLM) evaluated the impact of the dif-
ferent layouts on student academic outcomes. Hedge’s (g) effect size
calculations evaluated the effect of the various layouts on the six in-
tervention classes (II) while moderating the influence of the classroom
teachers, class composition, and student cognitive ability. The addition
of the HLM, to ascertain the variance in student academic results, ac-
counted for the potential influence of assessment, curriculum, the no-
velty effect of the intervention and the short-term nature of the within-
group design. A two-level (level-1 and -2) HLM applied an ordinary
least square (OLS) regression-based analysis to determine the effect of
key independent variables, as suggested by Woltman, Feldstain,
MacKay, and Rocchi, (2012), on student English, Humanities and
Mathematics achievement. The level-2 variables consisted of class
grouping, classroom layout (ILE and traditional), subject, and teacher.
The level-1 variables consisted of student cognitive ability (AAS in-
strument) and student grade, the latter the dependent variable of in-
terest in the HLM. The HLM process was selected due to its prior ap-
plication by Tanner (2008) who determined a 2%–7% of the variance in
academic achievement due to a school’s physical learning environment.
Even though the Tanner study was not explicitly focused on the com-
parison of traditional and ILEs, it evaluated the impact of the physical
design features comparable to those found in ILEs on student academic
progress incomparable subject matter domains (reading, vocabulary,
math, social studies).
4.1. Evaluation of the spatial intervention on student attitudes about their
learning
Student learning experiences.
The within-group analysis of the Scale A ascertained if students
noted changes in those learning experiences often associated with a
more student-centred learning approach (i.e., active learning, colla-
boration, creativity, and personalisation). RM-ANOVAs of classes 7.2,
8.1 9.1, 9.2 and 9.3 indicated a significant time effect (Table 4). The
post hoc visual analysis of classes 9.1, 9.2 and 9.3 meet the criterion of
statistically significant effects (SS), with a sustained effect between the
traditional to ILE layouts. However, the visual analysis revealed a de-
layed statistical effect for classes 7.1 and 7.2. While the visual analysis
of class 8.1 highlighted unstable responses during the intervention
phase that resulted in overlapping CIs at the second post-intervention
measure.
For Scale B, the RM-ANOVA and post hoc comparisons (Table 4)
indicated a significant time effect, which was attributed to the timing of
the ILE intervention, for all but one class. The process indicated sus-
tained statistically significant time effects for the classes 7.1, 8.1, 9.1,
9.2, and 9.3. In a similar vein to the post hoc visual analysis for Scale A,
class 7.2 indicated a delayed effect, but this time at the third post-in-
tervention measure (see Fig. 4).
In response to the first research question, the analysis of Scales A
and B showed how student attitudes toward their learning experiences
differed between the two layouts. Directly after the intervention, four of
the six classes indicated that their teachers changed some aspect of their
practice. There was an increased agreement amongst these students that
their learning experiences were somehow more active, collaborative,
personalised and interactive. Importantly, the analysis of Scales A and B
presented conclusions that were similar to the earlier studies at this site,
with a similar student demographic (Imms and Byers, 2016; Byers and
Imms, 2014). The replication across several studies does begin to build
an argument about the potential impact of different spatial layouts on

Table 3
Post Hoc Reliability Analysis through Cronbach’s Alphas for Individual Survey Scales.
Survey Domain Items Pre-ILE Intervention Post-ILE Intervention

Pre 1 Pre 2 Pre 3 Post 1 Post 2 Post 3

Student-centred Learning experiences 5 .783 .784 .808 .801 .827 .832

Spatial Effect on Student-centred Learning 3 .829 .848 .882 .864 .859 .851
Student Motivational Values 5 .792 .815 .810 .805 .800 .833
Student Motivational Expectations 4 .821 .839 .848 .800 .823 .831
Spatial Effect on Engagement 2 .786 .769 .792 .785 .803 .806

172
T. Byers et al. Studies in Educational Evaluation 58 (2018) 167–177

Fig. 4. Post-hoc visual analysis to discern statistically significant (SS - Class 7.1 on the left) and not statistically significant (NS - Class 7.2 on the right) changes in
survey responses.

student attitudes to their learning.
The impact of changing spatial layouts on student motivation.
The analysis of the LPTS Scales C (motivational values) and D
(motivational expectancies) responded to the second research question,
which focused on the possible effect of different spatial layouts on
student motivation. The RM-ANOVA and post hoc analysis of responses
to these scales (Table 5) found that the transition did not correlate with
statistically significant changes in attitudes. While general improve-
ments in attitudes to both were found; none meet the criteria for the
statistically significant difference. The visual analysis indicated that
student self-assessment of their motivation for learning throughout the
baseline and intervention phases was consistently favourable, with low
variance within classes.
4.2. The effect of changing classroom layout on student engagement
The analysis of student responses to Scale E measured attitudes to
their cognitive and emotional engagement to address the third research
question. The RM-ANOVA and post hoc comparisons (Table 6) in-
dicated all classes except class 7.2 indicated a significant time effect
attributed to the spatial intervention. Like the analysis of Scales A and
B, even though the RM-ANOVA indicated a significant effect for class
7.2, the visual analysis revealed a delayed statistical effect that oc-
curred after the second post-intervention data point.
Analysis of Scale E suggested that while in the ILE, this sample re-
ported a higher assessment of their attitudes about their cognitive and
emotional engagement than in the traditional layout. Students in this
sample, like those of the earlier Byers and Imms (2014) and Imms and
Byers (2016) studies, indicated a significantly higher assessment of
their engagement in learning. The visual analysis revealed a higher
agreement amongst the classes during the ILE phase, compared to the
traditional phase, through relatively small variance in the CIs. To-
gether, these results present a reliable indication of how students as-
sessed an ILE as their preferred learning environment (behavioural
engagement) and associated it with a greater interest in their learning
(cognitive engagement).
4.3. Longitudinal impact of different learning spaces on student attitudes
4.3.1. The impact of classroom type on student learning experiences
The between-group RM-ANOVA and post hoc comparisons (Table 7)
of the cognitively matched control classes to Scales A and B revealed
similar trends to the within-group intervention class analysis. Analysis
of Scale A revealed a positive and stable statistically significant spatial
effect (learning space x time) between four of the between-group
comparisons. For Scale B, again four of the between-group comparisons
resulted in a statistically and sustained significant time effect. The be-
tween-group RM-ANOVA of classes 7.4 (IC) vs. 7.6 (TC) (Scales A and
B), 8.2 (IC) vs. 8.6 (TC) (Scale A) and 8.4 (IC) vs. 8.7 (TC) (Scale B)
indicated a significant statistical effect. However, the post-hoc visual
analysis revealed an inconsistent mean trend or volatility in CIs of one
or both groups’ responses across the school year. The resulting overlap
in CIs (p > 0.05) at various points over the year were deemed not to
reflect a reliable and sustained statistically significant difference be-
tween these classes.
The similarities of the between-group analysis of Scales A and B,
replicating those of the intervention within-group analysis, suggest that
the different classroom spaces contributed (to some degree) to these

Table 4
Summary of Student Attitudinal Changes of their Learning Experiences through the Transition from Traditional Classroom to ILE.
Class Ability Learning Experiences Spatial Effect on Experiences

df F η p Visual df F η p Visual

7.1 Mixed 2.62 2.91 .19 .051 NSa 5 11.93 .48 < 0.001 SSb
7.2 Mixed 5 8.37 .27 .008 NS 5 6.12 .22 < 0.001 NS
8.1 Mixed 2.90 5.47 .22 .002 NS 5 14.25 .42 < 0.001 SS
9.1 High 5 11.19 .38 < .001 SS 5 17.46 .50 < 0.001 SS
9.2 Mixed 5 10.66 .37 < .001 SS 3.40 18.59 .49 < 0.001 SS
9.3 Mixed 2.85 10.36 .35 < .001 SS 3.49 12.88 .40 < 0.001 SS

a b
Note. Non-statistically significant visual effect. Statistically significant visual effect.

173
T. Byers et al. Studies in Educational Evaluation 58 (2018) 167–177

Table 5
Summary of Student Attitudinal Changes of their Learning Experiences through the Transition from Traditional Classroom to ILE.
Class Ability Motivational Values Motivational Expectancies

df F η p Visual df F η p Visual

a
7.1 Mixed 3.37 0.36 .03 0.874 NS 2.86 0.65 .03 0.665 NS
7.2 Mixed 5 2.49 .10 0.035 NS 5 4.78 .18 0.001 NS
8.1 Mixed 5 2.96 .13 0.016 NS 2.89 5.21 .21 0.003 NS
9.1 High 5 3.36 .16 0.008 NS 2.73 5.79 .24 0.002 NS
9.2 Mixed 3.11 1.62 .08 0.163 NS 5 4.20 .19 0.002 NS
9.3 Mixed 3.17 1.31 .07 0.266 NS 3.46 0.98 .05 0.434 NS

a
Note. Non-statistically significant visual effect.

Table 6 statistical difference. The similarities that have occurred in the be-
Summary of Changes in Student Perceptions of their levels of Engagement tween- and within-group comparisons would suggest that the different
through the Spatial Transformation from Traditional Classroom to ILE. spatial layouts had little effect on student motivation.
Class Ability Engagement in Learning
4.3.3. The longer-term effect of different spaces on student engagement
df F η p Visual The analysis indicated that in all between-group comparisons, each
7.1 Mixed 3.07 15.01 .41 < 0.001 SSa ILE class reported a higher level of cognitive and emotional engagement
7.2 Mixed 3.16 6.91 .35 0.002 NSb in their learning than their peers in a traditional layout. The RM-
8.1 Mixed 5 18.03 .47 < 0.001 SS ANOVAs and post-hoc calculations (Table 9) indicated a clear and
9.1 High 5 22.77 .56 < 0.001 SS consistent statistically significant difference, corroborated by high
9.2 Mixed 5 19.02 .51 < 0.001 SS
partial eta squared (η2) values, between all ILE and traditional classes.
9.3 Mixed 3.36 13.81 .42 < 0.001 SS
The corroboration of the between- and within-group analysis within
Note. a Statistically significant visual effect. b Non-statistically significant visual this study presents a compelling picture of how the students in an ILE
effect. viewed their engagement in their learning. Furthermore, the similarity
in findings now across three studies at this site (see the earlier Byers and
differences in student attitudes. Even though the cognitively matched Imms, 2014 and Imms and Byers, 2016 studies) suggests that different
classes covered comparable curriculum material and assessment items, spatial layouts can have a significant impact on student assessment of
the classes in an ILE responded more favourably to questions focused on their engagement in learning, with a favourable evaluation in an ILE.
active, collaborative, creative and personalised learning experiences
than their peers in a traditional layout. Students who occupied an ILE 4.4. The effect of different classroom spaces on student academic results
consistently reported a more favourable assessment of their specific
learning environment having a positive effect on their attitudes to their 4.4.1. Impact of the transitions from traditional classrooms to ILEs on
learning experiences. Furthermore, the post-hoc analysis revealed that academic performance
as the year progressed, the differences between all but one of the be- The application of Hedge’s (g) compared student academic perfor-
tween-group comparisons became more pronounced. The replicated mance of the six intervention classes through the baseline (first seme-
nature of the between- and within-group comparisons to both Scales A ster of the school year in a traditional classroom) and intervention
and B strengthens the argument that different spaces can affect student (second semester of the school year in ILE) phases (see Table 10). Like
learning experiences. the analysis of the LPTS survey, the application of Hedge’s (g) suited the
class-based unit of analysis (n < 30). The calculation of effect sizes
4.3.2. The impact of different spaces on student motivation showed that all classes academic results improved through the spatial
For the motivation scales (C and D), the results of the between- transition, with the most pronounced gain in Mathematics. Interest-
group analysis were comparable to those obtained for within-group ingly, the effect sizes calculated here where similar to those presented
comparisons. In the ILE, all classes gave a higher assessment of their in the earlier study by the Authors at the same site (2014a) (averaged
motivation expectancies and values than in a traditional classroom. The Mathematics g = .41, English g = .40 respectively).
RM-ANOVA and post hoc comparisons (Table 8) indicated a statistically
significant difference between the high-ability classes 8.2 (IC) and 8.6 4.4.2. Comparative analysis of traditional classrooms and ILE layouts on
(TC). However, like the earlier within-group analysis the relatively high longer-term academic performance
and homogeneous assessment of motivation by both control groups The HLM analysis investigated the degree of variance in student
meant that there was limited scope to achieve a clear and consistent academic performance (levels of achievement on an A+ to E− scale)

Table 7
Summary of Between-group Comparison of the ILE and Traditional Control Class Perceptions of their Learning Experiences.
Class Ability Learning Experiences Spatial Effect on Experiences

df F η p Visual df F η p Visual

7.3 vs. 7.5 High 1 28.38 .50 < .001 SSa 1 24.53 .42 < .001 SS
7.4 vs. 7.6 Mixed 1 4.83 .15 .036 NSb 1 11.05 .28 .002 NS
8.2 vs. 8.6 High 1 5.47 .22 .002 NS 1 125.35 .70 < .001 SS
8.3 vs. 8.7 Mixed 1 164.26 .75 < .001 SS 1 42.04 .55 < .001 SS
8.4 vs. 8.7 Mixed 1 21.55 .38 < .001 SS 1 19.21 .32 < .001 NS
8.5 vs. 8.7 Mixed 1 26.70 .51 < .001 SS 1 48.12 .59 < 0.001 SS

a b
Note. Statistically significant visual effect. Non-statistically significant visual effect.

174
T. Byers et al. Studies in Educational Evaluation 58 (2018) 167–177

Table 8
Summary of Between-group Comparison of the ILE and Traditional Control Class Perceptions of their Motivational Expectancies and Values.
Class Ability Motivational Values Motivational Expectancies

df F η p Visual df F η p Visual

a
7.3 vs. 7.5 High 1 4.88 .22 .037 NS 1 12.05 .34 .002 NS
7.4 vs. 7.6 Mixed 1 16.06 .32 .036 NS 1 8.46 .21 .011 NS
8.2 vs. 8.6 High 1 39.59 .42 < .001 SSb 1 41.07 .44 < .001 SS
8.3 vs. 8.7 Mixed 1 11.48 .25 .002 NS 1 10.10 .22 .003 NS
8.4 vs. 8.7 Mixed 1 16.43 .29 < .001 NS 1 13.55 .25 .001 NS
8.5 vs. 8.7 Mixed 1 11.82 .26 .002 NS 1 12.97 .28 .001 NS

a b
Note. Non-statistically significant visual effect. Statistically significant visual effect.

Table 9 layouts resulted in a statistically significant effect on academic results

Summary of Between-group Comparison of the ILE and Traditional Control in English [R2 =0.02, F(1, 127) = 4.73, p = 0.032] Humanities [R2 =
Class Perceptions of their Levels of Engagement. 0.03, F(1, 127) = 5.79, p = 0.018] and Mathematics [R2 = 0.09, F(1,
Class Ability Engagement in Learning 127) = 27.83, p < 0.001].
To evaluate the impact of space on the relationship between the
df F η p Visual identified independent variables (class grouping, classroom space,
cognitive ability, and teacher) and grades, a linear regression model of
7.3 vs. 7.5 High 1 43.59 .66 < .001 SSa
7.4 vs. 7.6 Mixed 1 126.87 .82 < .001 SS the following form was employed:
8.2 vs. 8.6 High 1 197.27 .78 < .001 SS
8.3 vs. 8.7 Mixed 1 71.78 .72 < .001 SS
GRADE = β 0 + β1COG + β 2TEACH + β 3CLASS + β 4SPACE + ε
8.4 vs. 8.7 Mixed 1 55.42 .61 < .001 SS
8.5 vs. 8.7 Mixed 1 121.28 .78 < .001 SS
Keeping with the quasi-experimental design, the calculation of un-
Note. a
Statistically significant visual effect. standardized (B) and standardised (β) regression coefficients, and
squared semi-partial correlations (sr2) for each predicator for the
Table 10 second step are reported for each subject in Years 7–9 (Table 11).
Within-group Hedge’s (g) Effect Size Calculation of the Intervention Class Seven out of the nine HLMs indicated that the various classroom
Academic Performance in English, Humanities, and Mathematics. spaces did account for a significant effect on student grades. The sta-
tistically significant variances in student grades ranged from 2% to
Class Ability English Humanities Mathematics
11%. Interestingly, this range was somewhat similar to the range
7.1 Mixed +0.55 +0.24 +0.40 identified by Tanner (2008). Interestingly, the influence of the different
7.2 Mixed +0.56 +0.11 +0.73 learning spaces, across this sample, contributed the second highest (to
8.1 Mixed +0.26 +0.51 +0.66
student cognitive ability) variance in student academic results. The
9.1 High +0.41 +0.33 +1.19
9.2 Mixed +0.44 +0.47 +1.03 classroom layout type had an equal to or greater impact than the dif-
9.3 Mixed +0.53 +0.12 +0.88 ferent classroom teachers or class groupings. It is acknowledged that
the random allocations of teachers across the study and ability group-
ings of some classes may have affected the interaction between the
attributed to the different spatial layout and other non-school (student grade variance attributed to class groupings and teacher (i.e., Year 8
cognitive ability) and school (teacher and class composition) factors in Mathematics and Year 9 English had different teachers for the eight
the second semester of the school year. The use of second-semester classes). Besides the acknowledged influence of cognitive ability on
academic data, post the spatial intervention for the II classes, sought to student academic achievement (Rohde & Thompson, 2007), the find-
control the potential influence of student maturation. Before the HLM, ings presented here replicated those of the earlier studies with the
several assumptions were assessed. First, box and stem plots indicated different layouts having a notable variance on academic results, espe-
that each variable in the regression was normally distributed and free cially in Mathematics and to a lesser extent English (Byers and Imms,
from univariate outliers. Second, inspection of the plots of the normal 2014).
probability of standardised residuals against standardised predicted
values met the assumptions of homoscedasticity, linearity, and nor-
mality of residuals. Finally, Mahalanobis distances (a measure of var- 5. Conclusion
iance of each variable and the covariance between variables to identify
outliers) did not exceed the critical χ2 for df =2 of 13.82 and df =3 of This study presented a set of empirical methods able to evaluate the
16.27 for any subjects or year levels. impact of different learning spaces on teaching and learning. The quasi-
As expected in all year-levels and subjects the combinations of class experimental approach facilitated by an SSRD sought to isolate the link
grouping, cognitive ability, and teacher (step 1) accounted for a sta- between the different classroom layouts and student attitudes related to
tistically significant variability in the learning outcomes. When ad- their learning. In this process, this study provided further empirical
justed for the effect of the different classroom layouts (step 2), this evidence that goes some way to better understand a possible relation-
resulted in the statistically significant adjusted R2 values for students’ ship between different learning spaces and teaching and learning.
grades in English [R2 =0.11 F(1, 79) = 17.03, p < 0.001], Humanities However, the evidence presented here, in association with the earlier
[R2 = 0.07, F(1, 79) = 9.10, p = 0.003], and Mathematics [R2 =0.04, studies at this site, does assert that this relationship is not causal and
F(1, 79) = 6.93, p = 0.029]. For Year 8, only the analysis of goes beyond a simple spatial change from traditional classrooms to
Mathematics indicated that the adjustment for the effect of the different ILEs.
classroom layouts (step 2) resulted in a statistically significant effect in In this sample, the positive and at times statistically significant ef-
student grades [R2 = 0.04 F(1, 157) = 16.60 p < 0.001]. Finally, in fects suggest that the difference in spatial layouts affected student at-
Year 9 the analysis adjusted for the effect of the different classroom titudes to their learning experiences and engagement. While in an ILE,
students identified a wider array of active learning experiences and

175
T. Byers et al. Studies in Educational Evaluation 58 (2018) 167–177

Table 11
Step 2 of Single-Level Regression Results for Year 7, 8 and 9 Academic Results in English, Humanities, and Mathematics.
Year Variable English Humanities Mathematics

B β sr2
B β sr 2
B β sr2

Constant 7.80 (.60) 8.06 (.71) 6.87 (.78)

Ability .55 (.10)* .67 .19 .58 (.12)* .47 .18 .87 (.14)* .62 .25
7 Teacher −.42 (.29) −.33 .02 −.12 (.20) −.06 .00 −.31 (.33) −.21 .01
Class −1.14(.31)* .00 .09 −1.24(.34)* −1.14 .10 −.92 (.52) −.71 .02
Classroom 2.65(.64)* .64 .11 2.23(.74)* .99 .07 1.78(.79)* .65 .04
Constant 7.73 (.65) 7.79 (.61) 3.47 (.56)
Ability .37(.07)* .38 .11 .55 (.08)* .53 .20 1.09(.07)* .81 .55
8 Teacher −.56(.14)* −.63 .08 −.61 (.18)* −.52 .03 −.49 (.07)* −.17 .03
Class .32(.10)* .46 .04 −.42(.14)* −.50 03 .00 (.00) .00 .00
Classroom .29 (.18) .18 .01 .31 (.14) .11 .01 .57(.22)* −.18 .04
Constant 7.27 (.59) 6.44 (.70) 4.25 (.75)
Ability .69 (.08)* .56 .33 .74 (.09)* .60 .37 1.10(.09)* .72 .40
9 Teacher −.28(.05)* −.35 .09 −.36 (.45) −.31 .01 −.39 (.12)* −.31 .03
Class .00 (.00) .00 .00 .16 (.30) .21 .00 .17 (.11) .15 .00
Classroom .54(.25)* .15 .02 .80 (.33)* .21 .03 1.72 (.33)* ..32 .09

Note. Unstandardized Coefficients (B) reported standard errors in parentheses. Note. * p < .05.

collaborative learning modalities than their peers who remained in a
traditional layout. However, the ILE by themselves are not the agents of
change, with a notable difference across the sample. The analysis re-
vealed not all classes in an ILE (for example ILE intervention Class 7.2)
indicated a significant and sustained change in the nature of learning
experiences when compared to their time in the traditional layout. As a
consequence, for this sample, the occupation of an ILE supported those
teachers who were able and willing to align better the affordances of
the physical learning environment to a wider array of pedagogies.
This study has shown that different learning spaces can affect tea-
cher pedagogies, which inturn impact, student academic outcomes. By
accounting for the effects of various confounding variables (i.e., class
groupings, cognitive ability and teacher), it has shown that the different
classroom layouts influenced variance in student grades. Students who
occupied an ILE tended to outperform their like-ability peers in a tra-
ditional design. As expected, the HLM indicated that student cognitive
ability had a more significant impact on the variance in student grades
than the spatial layout, for this sample. However, the HLM for the
sample indicated the somewhat similar variance attributed to the tea-
cher and the different learning. Furthermore, the range of significant
variances in student grades was similar to the range identified by
Tanner (2008).
The analysis reinforced the mediating role of the classroom teacher
within the interplay that exists between pedagogy and space.
Comparisons between LPTS survey and academic achievement analysis
suggests that the most pronounced statistically significant effects oc-
curred in classes with teachers who were more able to use the addi-
tional affordances of the ILE for pedagogical gain. This phenomenon
supports the notion of teacher environmental competency as suggested by
both Lackney (2008) and Steele (1980). Both suggested that teacher
environmental competency underpins their capacity to navigate and
evolve their practices to utilise the affordances of the new spaces for
pedagogical gain. Future studies will evaluate the micro changes to
teaching and learning associated with the occupation of different
learning spaces to better understand teacher environmental compe-
tency. This is important, as such support is not widely available. It is
speculated that better teacher support would enhance their ability and
understanding of the use of space to better utilise the available affor-
dances of any spatial type (ILE or traditional) for pedagogical gain.
In conclusion, even though the evidence does suggest a tangible
link, it would be naïve to say that it is causal. In future studies, much
deeper analysis of quantitative data from multiple sites will better de-
termine how different learning environments affect teaching and
learning. These studies will also investigate the contribution of school
(collective socioeconomic status, leadership, location and type) factors
that this study was unable to investigate. At the same time, such a study
will better discern the mediating influence of the classroom teacher
with a greater and more diverse teacher sample. Finally, multiple site
analysis will investigate the potential contribution of key school factors
often identified in the literature (i.e. leadership and school type) on
teacher transition from traditional classrooms to ILEs.
References
Aldridge, J. M., Fraser, B. J., Bell, L., & Dorman, J. P. (2012). Using a new learning
environment questionnaire for reflection in teacher action research. Journal of Science
Teacher Education, 23(3), 259–290. https://doi.org/10.1007/s10972-012-9268-1.
Baguley, T. (2012). Calculating and graphing within-subject confidence intervals for
ANOVA. Behavior Research Methods, 44(1), 158–175. https://doi.org/10.3758/
s13428-011-0123-7.
Barrett, P. S., Zhang, Y., Moffat, J., & Kobbacy, K. (2013). A holistic, multi-level analysis
identifying the impact of classroom design on pupils’ learning. Building and
Environment, 59, 678–689. https://doi.org/10.1016/j.buildenv.2012.09.016.
Benade, L. (2016). Is the classroom obsolete in the twenty-first century? Educational
Philosophy and Theory, 1–12. https://doi.org/10.1080/00131857.2016.1269631.
Blackmore, J., Bateman, D., O’Mara, J., & Loughlin, J. (2011). Research into the connection
between built learning spaces and student outcomes: Literature review. Melbourne:
Victorian Department of Education and Early Childhood Development. Retrieved
from http://www.eduweb.vic.gov.au/edulibrary/public/publ/research/publ/
blackmore_learning_spaces.pdf.
Brooks, D. C. (2011). Space matters: The impact of formal learning environments on
student learning. British Journal of Educational Technology, 42(5), 719–726. https://
doi.org/10.1111/j.1467-8535.2010.01098.x.
Byers, T. (2016). Evaluating the effects of different classroom spaces on teaching and learning.
(Doctor of Philosophy). Melbourne: The University of Melbourne.
Byers, T. (2017). What does teaching and learning look like in different classroom en-
vironments? Paper presented at the 2017 transitions research symposium: inhabiting in-
novative learning environments.
Byers, T., Hartnell-Young, E., & Imms, W. (2018). Empirical evaluation of different
classroom spaces on student perceptions of the use and effectiveness of 1 to 1 tech-
nology. British Journal of Educational Technology, 49(1), 153–164. https://doi.org/10.
1111/bjet.12518.
Byers, T., & Imms, W. (2014). Making the space for space: The effect of the classroom
layout on teacher and student usage and perception of one-to-one technology. Paper
presented at the 26th Australian computers in education conference.
Byers, T., & Imms, W. (2016). Evaluating the change in space in a technology-enabled
primary years setting. In K. Fisher (Ed.). The translation design of schools: An evidence
based approach to aligning pedagogy and learning environment design (pp. 215–236). The
Netherlands: Sense.
Byers, T., Imms, W., & Hartnell-Young, E. (2014). Making the case for space: The effect of
learning spaces on teaching and learning. Curriculum and Teaching, 29(1), 5–19.
https://doi.org/10.7459/ct/29.1.02.
CABE (2005). Picturing school design. A visual guide to secondary school buildings and their
surroundings using the design quality indicator for schools. London: Commission for
Architecture and the Built Environment.
Chandler, W. L. (2009). A teacher space or a learner place? Reconsidering the classroom
environment. International Journal of Learning, 16(9), 261–267. http://www.
Learning-Journal.com.
Cleveland, B. W., & Fisher, K. (2014). The evaluation of physical learning environments:
A critical review of the literature. Learning Environments Research, 17(1), 1–28.
https://doi.org/10.1007/s10984-013-9149-3.

176
T. Byers et al.
Dorman, J. P., Aldridge, J. M., & Fraser, B. J. (2006). Using students’ assessment of
classroom environment to develop a typology of secondary school classrooms.
International Education Journal, 7(7), 906–915 doi:http://iej.com.au.
Dorman, J. P., & Fraser, B. J. (2009). Psychosocial environment and affective outcomes in
technology-rich classrooms: testing a causal model. Social Psychology of Education,
12(1), 77–99. https://doi.org/10.1007/s11218-008-9069-8.
Dovey, K., & Fisher, K. (2014). Designing for adaptation: The school as socio-spatial as-
semblage. The Journal of Architecture, 19(1), 43–63. https://doi.org/10.1080/
13602365.2014.882376.
Dumont, H., & Istance, D. (Eds.). (2010). Analysing and designing learning environments for
the 21st century. Paris: Organisation for Economic Cooperation and Development.
Fraser, B. J. (1982). Differences between student and teacher perceptions of actual and
preferred classroom learning environment. Educational Evaluation and Policy Analysis,
4(4), 511–519.
Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School engagement: Potential of
the concept, state of the evidence. Review of Educational Research, 74(1), 59–109.
https://doi.org/10.3102/00346543074001059.
Gersten, R., Fuchs, L. S., Compton, D., Coyne, M., Greenwood, C., & Innocenti, M. S.
(2005). Quality indicators for goup experimental and quasi-experimental research in
special education. Exceptional Children, 71(2), 149–164. https://doi.org/10.1177/
001440290507100202.
Ghasemi, A., & Zahediasl, S. (2012). Normality tests for statistical analysis: A guide for
non-statisticians. International Journal of Endocrinology and Metabolism, 10(2),
486–489. https://doi.org/10.5812/ijem.3505.
Gislason, N. (2010). Architectural design and the learning environment: A framework for
school design research. Learning Environments Research, 13(2), 127–145. https://doi.
org/10.1007/s10984-010-9071-x.
Gliem, J. A., & Gliem, R. R. (2003). Calculating, interpreting, and reporting cronbach’s
alpha reliability coefficient for likert-type scales. Paper Presented at the Midwest
Research-to-Practice Conference in Adult, Continuing, and Community Education.
Hattie, J. A. C., & Donoghue, G. M. (2016). Learning strategies: A synthesis and con-
ceptual model. Science of Learning, 1, 1–13. https://doi.org/10.1038/npjscilearn.
2016.13.
Higgins, S., Hall, E., Wall, K., Woolner, P., & McCaughey, C. (2005). The impact of school
environments: a literature review. Newcastle: University of Newcastle.
Imms, W., & Byers, T. (2016). Impact of classroom design on teacher pedagogy and
student engagement and performance in mathematics. Learning Environments
Research, 19(2), 1–14. https://doi.org/10.1007/s10984-016-9210-0.
Imms, W., Mahat, M., Byers, T., & Murphy, D. (2017). Type and use of innovative learning
environments in Australasian schools ILETC survey 1. Retrieved from Melbourne.
Lackney, J. A. (2008). Teacher environmental competence in elementary school en-
vironments. Children, Youth and Environments, 18(2), 133–159 doi:www.color-
ado.edu/journals/cye.
Lambert, N. N., & McCombs, B. L. (Eds.). (1988). How students learn: Reforming schools
through learner-centered education. Washington: APA Books.
McGregor, J. (2004). Spatiality and the place of the material in schools. Pedagogy, Culture
& Society, 12(3), 347–372. https://doi.org/10.1080/14681360400200207.
177
Studies in Educational Evaluation 58 (2018) 167–177
Monahan, T. (2002). Flexible space & built pedagogy: Emerging IT embodiments. Inventio,
4(1), 1–19 doi:http://www.torinmonahan.com/papers/Inventio.html.
Mulcahy, D. (2016). Policy matters: de/re/territorialising spaces of learning in Victorian
government schools. Journal of Education Policy, 31(1), 81–97. https://doi.org/10.
1080/02680939.2015.1099077.
Mulcahy, D., & Morrison, C. (2017). Re/assembling ‘innovative’ learning environments:
Affective practice and its politics. Educational Philosophy and Theory, 1–10. https://
doi.org/10.1080/00131857.2016.1278354.
OECD (2013). Innovative learning environments. Paris: Organisation for Economic
Cooperation and Development.
Painter, S., Fournier, J., Grape, C., Grummon, P., Morelli, J., Whitmer, S., & Cevetello, J.
(2013). Research on learning space design: Present state, future directions. Ann Arbor:
Society for College and University Planning.
Pintrich, P. R., & De Groot, E. V. (1990). Motivational and self-regulated learning com-
ponents of classroom academic performance. Journal of Educational Psychology, 82(1),
33–40. https://doi.org/10.1037/0022-0663.82.1.33.
Reynard, R. (2009). Designing learning spaces for instruction, not control. Campus
Technology. Retrieved from http://campustechnology.com/articles/2009/04/29/
designing-learning-spaces-for-instruction-not-control.aspx.
Rohde, T. E., & Thompson, L. A. (2007). Predicting academic achievement with cognitive
ability. Intelligence, 35(1), 83–92. https://doi.org/10.1016/j.intell.2006.05.004.
Ryan, A. M., & Patrick, H. (2001). The classroom social environment and changes in
adolescents’ motivation and engagement during middle school. American Educational
Research Journal, 38(2), 437–460.
Sanoff, H. (2001). School building assessment methods. Washington DC: National
Clearinghouse for Educational Facilities..
Soccio, P. (2014). Using the EduTool:IEQ to evaluate IEQ performance inside classrooms.
2014, 28 Oct 2014 - 30 Oct 2014 Paper Presented at the World Sustainable Buildings.
Steele, F. (1980). Defining and developing environmental competence. In C. P. Alderfer, &
C. L. Cooper (Eds.). Advances in experimental social processes 2 (pp. 225–244). New
York: Wiley.
Tamim, R. M., Lowerison, G., Schmid, R. F., Bernard, R. M., & Abrami, P. C. (2011). A
multi-year investigation of the relationship between pedagogy, computer use and
course effectiveness in postsecondary education. Journal of Computing in Higher
Education, 23(1), 1–14. https://doi.org/10.1007/s12528-010-9041-4.
Tanner, C. K. (2008). Explaining relationships among student outcomes and the school’s
physical environment. Journal of Advanced Academics, 19(3), 444–471.
Tanner, C. K. (2009). Effects of school design on student outcomes. Journal of Educational
Administration, 47(3), 55–73. https://doi.org/10.1108/09578230910955809.
Wall, C. (2009). Building the education revolution: National coordinator’s implementation
report (February–September 2009)Canberra: Australian Federal Government.
Woltman, H., Feldstain, A., MacKay, J. C., & Rocchi, M. (2012). An introduction to
hierarchical linear modeling. Tutorials in Quantitative Methods for Psychology, 8(1),
52–69 doi: http://www.tqmp.org/.
Zandvliet, D. B., & Fraser, B. J. (2004). Learning environments in information and
communications technology classrooms. Technology, Pedagogy and Education, 13(1),
97–123. https://doi.org/10.1080/14759390400200175.