Abstract
Whilst beneficial, the use of evidence to improve teaching and learning in schools is proving difficult to achieve in practice. The purpose of this paper is to shed new light on this issue by examining the applicability of a model of rational behaviour as relates to the notion of evidence-informed practice (EIP). Specifically, exploring the question: “if EIP is rational behavior, why are not all teachers engaged in it?”, the authors examine whether the beliefs and perspectives of teachers in relation to EIP, align with their evidence-use behaviours. The authors then assess what factors prevent teachers/schools from engaging in EIP. Using a Gradient Boosted Tree predictive model to analyse data from a survey of 696 practitioners in 79 schools, the findings suggest that school leaders need to promote the vision for evidence-use, illustrate its effectiveness, and establish learning environments for evidence-based conversations.
Introduction: The Evidence-Informed Practice Paradox
Evidence-Informed Practice (EIP) has become a cornerstone of modern educational reform. The underlying logic, often viewed through the lens of Rational Choice Theory (RCT), suggests that if research identifies effective teaching methods, rational practitioners will adopt them to maximize student outcomes. However, a persistent gap remains between teachers' positive attitudes toward research and their actual classroom behavior. This paper investigates this "rationality gap," questioning whether the lack of evidence use is a personal failing or a rational response to organizational constraints.
Theoretical Framework: Optimal Rationality
To understand this discrepancy, the study introduces the concept of Optimal Rationality (OR). Unlike traditional economic models that assume individuals always act on the best available information, OR posits that human behavior is a negotiation between "practical rationality"—the immediate actions required to manage daily workloads—and "cultural rationality"—the norms, values, and expectations set by the school organization. In this framework, a teacher might rationally choose not to use evidence if the school culture does not actively support or reward it, even if they believe in the evidence's value.
Methodology and Data
The research employed a quantitative approach, utilizing a survey of 696 practitioners across 79 primary schools in England. The sample was predominantly female (84%), with a significant portion in leadership roles (42% subject leaders and 19% in senior leadership).
The researchers moved beyond standard linear analysis by employing a Gradient Boosted Tree (GBT) predictive model. GBT is an ensemble machine learning method that constructs multiple decision trees to identify complex, non-linear relationships between variables. The model categorized survey data into "cause" variables (e.g., leadership support, trust, resources) and "effect" variables (e.g., teacher beliefs and actual EIP behaviors). Additionally, survey responses were linked to national OfSTED inspection grades to assess the impact of external accountability.
Key Findings: The Belief-Behavior Paradox
The analysis revealed a stark contrast between belief and action. While 76% of teachers agreed on the value of research, only 50% supported school-wide policy changes based on that research. This suggests that EIP is often viewed as an individual professional "hobby" rather than a collective institutional requirement.
The Impact of Accountability
A significant finding was the "OfSTED Chill." Schools rated as "Requires Improvement" or "Inadequate" were the least likely to engage with research. In high-stakes environments, these schools often prioritize "quick fixes" to satisfy inspectors over the long-term investment required for evidence-based innovation.
Predictors of Evidence Use
The GBT model identified three primary drivers for increasing EIP adoption:
- Encouragement (0.221): Active, visible prompting from school leadership.
- R-conversations (0.149): Creating formal and informal spaces for staff to discuss research findings.
- R-strategies (0.113): Integrating research into school-wide improvement plans.
While trust in leadership was found to be important, its direct predictive weight was lower (approx. 0.05) than specific, active leadership interventions.
Conclusion and Policy Implications
The study concludes that EIP is a "fragile" practice. Without explicit prioritization from school leaders, teachers will rationally default to "safe," traditional behaviors to meet immediate accountability demands. For EIP to be meaningfully realized, it must move from being a personal preference to a school-wide commitment. Furthermore, the authors argue that until EIP is integrated into national accountability regimes—such as inspection criteria—it is unlikely to become a systemic reality in schools facing high-stakes pressure.
Summary updated