Measuring the impact of productive failure on nursing students' learning in healthcare simulation: A quasi-experimental study

Background: Previous research suggests that making errors in a non-threatening simulated environment can facilitate learning. Productive failure, which combines problem-solving tasks followed by instruction, enables students to learn from making mistakes. This teaching approach has demonstrated improved learning outcomes such as explanatory knowledge and transfer of knowledge compared to a direct instruction approach where students receive instruction prior to problem-solving tasks. However, no previous studies have examined the impact of productive failure on nursing students' learning in manikin-based simulation. Objective: To measure the impact of productive failure on nursing students' declarative knowledge, explanatory knowledge, and transfer of knowledge compared to a direct instruction approach in a paediatric closed head injury simulation. Methods: Second year undergraduate nursing students (n = 349) from one Australian university were invited to participate in the study. Consenting participants (n = 344) were randomised into two groups: productive failure and direct instruction. The intervention consisted of two paediatric closed head injury simulations separated by a simulation debrief. Knowledge tests were administered before and immediately after the simulation. Results: Data from 331 participants were analysed. The productive failure group outperformed the direct instruction group in the post-test (p < 0.001). Learning gains for participants in the productive failure group were significantly higher than the direct instruction group for both explanatory knowledge (p < 0.001) and the ability to apply learning to solve novel clinical problems (p < 0.001). The difference in the median scores for declarative knowledge was not significant (p = 0.096). Conclusion: This study demonstrated that a productive failure simulation that leads learners to make mistakes before receiving instruction can facilitate deeper levels of explanatory knowledge and enable the transfer of learning to new clinical situations. These results suggest the need for further exploration of pedagogies that foster learning from errors in simulation-based learning.