Transfer of learning: thinking well

Ensuring students have a strong enough grasp of chemistry to be able to apply their knowledge in unfamiliar contexts is a key goal for any chemistry teacher, and it’s an ability being tested more and more in exams. This skill is sometimes referred to as transfer of learning, and is a particular challenge in STEM subjects.

Traditional teaching methods often result in shallow learning, characterised by rote-memorisation and a dependence on algorithmic methods. This leads to students being unable to apply learning to complex or unfamiliar situations.

In a recent study, Melonie Teichert and her co-workers explored the relationship between particular thinking processes used by students and their ability to transfer learning to address an unfamiliar problem.

Laboratory work in this study was based on the Model-Observe-Reflect-Explain (MORE) pedagogy, which has previously been shown to improve transfer success compared with standard instruction.

Students first develop initial models of the phenomenon they’re studying based on ideas from macroscopic and molecular-level perspectives. They then conduct experiments that compare and contrast several examples of the phenomenon. Then, they reflect on their observations and how they align with their initial model. Finally, students refine their sub-microscopic models and explain their rationale for doing so.

A key feature of the MORE Thinking Frame is that the teacher prompts students to engage in metacognitive monitoring of their understanding of models – to think about how and why their conceptualisation evolves over time.

Reflective reasoning

In this case, students were first asked to describe their understanding of what happens when salt and sugar are added separately to water. They then carried out experiments designed to inform their models. These included adding different compounds to distilled water, recording observations and measuring electrical conductivities. ‘Reflection questions’ prompted students to refine their models and to explain their revisions.

To test the impact on students’ abilities to transfer learning, they read a passage explaining boiling point elevation and were then asked to predict, with reasoning, the effect of solutes on the boiling point of aqueous solutions.

Of the 28 students interviewed, 12 made accurate predictions with correct reasoning, while the others were unable to apply their knowledge to tackle the problem.

How correct the initial and final models were was not strongly associated with successful transfer of learning

The researchers found that nine of the ‘correct’ students used three specific processes strongly associated with successful transfer of learning. These are: constructing sub-microscopic models consistent with experimental evidence; engaging in accurate metacognitive monitoring of their evolving sub-microscopic models; and using evidence to justify sub-microscopic model refinements.

Interestingly, how correct the initial and final sub-microscopic models were was not strongly associated with successful transfer of learning. Despite the small sample size, this is valuable evidence that informs how we support students in developing and applying models.