Flexible learning spaces facilitate interaction, collaboration and behavioural engagement in secondary school

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0223607

Study aims and context

• Objective: Objectively measure and compare adolescent classroom behaviour between traditional classrooms and flexible learning spaces, and assess the effect of space and teaching approach on a range of classroom behaviours.

• Rationale: Globally, schools are shifting from traditional classrooms to flexible learning spaces to improve engagement and outcomes; empirical evidence on behaviour and learning outcomes in these spaces is limited.

• Key premise: Flexible spaces, when paired with student-centred pedagogy, may increase interaction, collaboration, and engagement, potentially benefiting learning outcomes in the long term.

Study design and setting

• Design: School-based cross-over trial conducted in nine NSW secondary schools.

• Participants: Grade Seven–Nine classes from 12 invited public schools; final valid data from sixty (n = 60) students observed.

• Sample characteristics:

  • Age: M = 13.2, ext{SD} = 1.0. (age in years)

  • Sex: ext{Female} = 27 ext{ (45%)}.

  • Socioeconomic background: SEIFA scores reported; mean SEIFA = 1013.31, ext{SD} = 73.61.

  • Ethnicity/cultural diversity: Represented >13 groups.

• Ethical approvals: University of Wollongong Human Ethics Research Committee (HE16/021) and NSW SERAP.

• Data collection period: Observations occurred in 2016–2017; two data collection timepoints per school, 1–2 weeks apart.

• Setting comparison: Each class experienced both environments on different days:

  • Traditional classrooms: standard single room (~M = 50 ext{ m}^2); desks in rows facing the front; teacher-led instruction.

  • Flexible learning spaces: open-plan spaces (~M = 83 ext{ m}^2) with varied furniture (grouped tables, standing workstations, ottomans, couches, writable tables/walls); little to no distinct front-of-class, resources around the room; space designed to support diverse learning approaches.

• Cross-over design details: Same group of students observed in both conditions; lesson content and teacher remained consistent across conditions; order of conditions counterbalanced by school.

Teacher development and pedagogical approaches

• Pre-study professional development: Tours of flexible-space schools; conferences; short courses on designing/teaching in flexible spaces; informal teacher networks; teachers identified as change agents with a shared vision for student-centred learning.

• Traditional classrooms pedagogy: Predominantly teacher-led; students work mostly individually; frequent teacher input/instruction; limited opportunities to stand/move or collaborate.

• Flexible learning spaces pedagogy: Student-centred, group-work focused; teachers provide initial instructions and ongoing guidance; students have autonomy to choose how to approach tasks, where to work, and which resources/furniture to use; spaces and furniture enable movement and group work.

Observational instrument and coding scheme

• Instrument: COS-5 Pianta-based classroom observation tool, adapted for this study.

• Observational categories (Table 1 in study):

  • Student level setting: ext{Whole class}, ext{ groups } (>6), ext{ groups } (\le 6), ext{ individual}

  • Mode of learning: ext{Teacher-led instruction}, ext{ working individually}, ext{ collaborating}, ext{ presentation-based}, ext{ reflective}, ext{ research-based}

  • Academic behaviour: ext{Actively engage}, ext{ passively engaged}, ext{ off-task verbal}, ext{ off-task motor}, ext{ off-task passive}

  • Interaction with peers: ext{Positive interaction}, ext{ negative interaction}, ext{ no interaction}

  • Interaction with teacher: ext{Positive interaction}, ext{ negative interaction}, ext{ no interaction}

  • Use of technology: ext{Active use}, ext{ passive use}, ext{ no use}

• Additions for this study: Included a separate measure for mode of learning and use of technology to capture elements specific to flexible spaces.

• Observers: One trained observer conducted all observations to maximize validity; two hours of training and practice with video recordings prior to data collection.

• Reliability/validity considerations: Coding scheme adapted to capture space- and pedagogy-specific behaviours; observer kept consistent coding across all observations.

Data collection procedure

• Observation protocol: Students observed during a 30-minute lesson, with 6 students observed per lesson and at 30-second intervals (total of 10 observations per student per lesson).

• Selection of students: Class list used; first three female and first three male students chosen for observation; same six students observed at the second data collection. If unavailable, next student of the same sex chosen. Students and teachers unaware of which students were observed.

• Lesson timing: Observation began ~10 minutes after lesson commenced, once settled.

• Observation context: Lessons spanned English, mathematics, geography, and history; data collection occurred in both environments for the same cohort.

• Data handling: Observations recorded as frequencies of codes, later converted to percentages of lesson time for each participant and period.

Statistical analysis

• Software: SPSS (Version 21) and STATA (Version 13).

• Primary analysis: Mixed-effects multilevel linear regression to compare traditional vs flexible spaces, accounting for within-child differences and clustering by school (random effect for school).

• Significance: p < 0.05 considered statistically significant.

• Effect size: Cohen’s d calculated from group means and standard deviations, using traditional classroom as the denominator; interpreted as small ≈ 0.2, medium ≈ 0.5, large ≈ 0.8.

• Formula reference (conceptual):

  • Cohen’s d: d = rac{M1 - M2}{SD_{pooled}}.

  • Note in study: effect sizes calculated with traditional classroom as the denominator.

Key results (summary)

• Sample and data availability:

  • Invitations: N = 243; Consent obtained for n = 203 (83%).

  • Observed students: n = 60 across nine schools; six students per school observed in each condition; two data points per school; six additional students replaced if absences occurred.

• Overall effects across behaviours (significant differences between traditional and flexible spaces the majority of times):

  • Whole-class time: Flexible spaces reduced time spent in whole-class settings relative to traditional rooms: d = -0.65, ext{ } p = 0.001.

  • Time in groups (>6): Increased in flexible spaces: d = 0.46, ext{ } p = 0.004.

  • Time in groups (≤6): Increased in flexible spaces: d = 0.61, ext{ } p = 0.001.

  • Individual work: Slightly less in flexible spaces; difference approached significance: d = -0.35, ext{ } p = 0.046.

• Modes of learning (four significant differences out of six):

  • Collaboration: Flexible spaces → significantly more time collaborating: d = 1.33, ext{ } p = 0.001. (very large effect)

  • Teacher-led instruction: Flexible spaces → less time: d = -0.75, ext{ } p = 0.001.

  • Working independently: Flexible spaces → less time: d = -0.79, ext{ } p = 0.001.

  • Presentation-based: Flexible spaces → more time: d = 0.65, ext{ } p = 0.001.

  • Reflective learning and Research-based: Non-significant differences (Reflective: d = 0.11, ext{ } p = 0.351; Research-based: d = -0.25, ext{ } p = 0.055).

• Engagement with lesson content:

  • Actively engaged: Flexible spaces → more time: d = 0.50, ext{ } p = 0.001.

  • Passively engaged: Not significantly different: d = -0.27, ext{ } p = 0.123.

  • Off-task verbal: Flexible spaces → less off-task verbal: d = -0.44, ext{ } p = 0.016.

  • Off-task motor: No significant difference: d = -0.13, ext{ } p = 0.367.

  • Off-task passive: No significant difference: d = -0.07, ext{ } p = 0.686.

• Interaction with peers:

  • Positive interaction: Flexible spaces → more positive peer interactions: d = 0.88, ext{ } p = 0.001.

  • No interaction with peers: Flexible spaces → less lack of interaction: d = -0.85, ext{ } p = 0.001.

  • Negative interaction: No significant difference: d = 0.21, ext{ } p = 0.173.

• Interaction with teachers:

  • Positive interaction with teacher: No significant difference: d = 0.03, ext{ } p = 0.820.

  • Negative interaction with teacher: No significant difference: d = -0.19, ext{ } p = 0.313.

  • No interaction with teacher: No significant difference: d = -0.03, ext{ } p = 0.825.

• Use of technology:

  • Active use: Not statistically significant: d = -0.18, ext{ } p = 0.070.

  • Passive use: Not statistically significant: d = -0.24, ext{ } p = 0.131.

  • No use: Significant increase in flexible spaces: d = 0.26, ext{ } p = 0.022.$$

  • Overall takeaway on tech use: Flexible spaces associated with less time using technology, driven by more frequent non-use during lessons.

Detailed interpretation of results

• Space-pedagogy interplay:

  • Flexible spaces with student-centred pedagogy reduce reliance on teacher-led instruction and whole-class formats, while promoting group work and collaboration.

  • This shift corresponds with higher active engagement and more positive peer interactions, suggesting enhanced engagement and collaborative learning processes in flexible spaces.

• Engagement continuum:

  • Engagement observed along a continuum (actively engaged to off-task). Flexible spaces increased time spent in active engagement and decreased verbal off-task behaviour, without a rise in negative interactions with peers.

• Technology use:

  • Although technology use reduced marginally in flexible spaces, the pattern was not uniformly significant across all technology-related codes; overall, non-use time increased in flexible spaces.

• Implications for learning outcomes:

  • Authors suggest that increased interaction, collaboration, and active engagement in flexible spaces may translate into better long-term learning outcomes, though direct causal links require further experimental study.

Implications for practice and policy

• Professional development: Emphasize environmental competence—teachers’ ability to maximize the affordances of flexible spaces and to implement student-centred pedagogy effectively.

• School design and investment: Flexible learning spaces should be accompanied by ongoing teacher support, suitable resources, and targeted professional development to realize potential benefits.

• Classroom management: Findings suggest autonomous, collaborative environments can still maintain constructive interactions and high engagement when supported by teachers who facilitate dialogue and content-focused interactions.

• Policy relevance: Supports national and international moves toward