Knowledge Retrieval

There have been significant changes in how we understand the nature of knowledge and what it means to know. It’s slowly transitioned to a state where knowledge is now perceived as located within particular ‘paradigms’ with their own rules, criteria, and symbolic representations. For example, computer science is a rising topic of study in secondary education, and it requires a firm understanding of logic along with the proper syntax to communicate. Similarly mathematics, physics, literature, or history, each gave their own particular modes of interpretation, rather than the traditional theory-free descriptions of the world. This view of knowledge reflects the values of the knowledge economy, where workers need to be able to think systematically, creatively, and critically. As this discussion transgresses into familiar territory, let’s shift the focus onto knowledge acquisition and retention.

The prevailing thoughts on human learning are guided by a few tacit assumptions. The first such assumption is that learning happens primarily when people encode knowledge and experiences. Our school systems are structured around teacher sharing information or concepts, which the students then store or encode in their minds. A related assumption is that retrieval of this information, defined as the active cue-driven process of reconstructing knowledge, measures the products of previous learning, but doesn’t itself produce learning. The idea behind this assumption is that measuring memory can’t change memory in the same way that the act of measuring a shape does not change the shape. However, Jeffrey D. Karpicke and Janell R. Blunt’s research shows that retrieval plays an active role in learning.

The figure above shows the result of an experiment where students were evaluated on science concepts after various study methods. In ‘Repeated Study’, students studied the text in four consecutive sessions. In the ‘Concept Mapping’ scenario, students studied the text in one session and then created a concept map of the material while viewing the text. Finally, the ‘Retrieval Practice’ method involved students studying the text for one session, then practical retrieval by recalling as much information on a free recall test. The total amount of learning time was equal to those in concept mapping and retrieval practices.
The results of this experiment prove two important things. Firstly, not only is retrieval a relevant metric in knowledge acquisition, it has the potential to exceed traditional study and encoding methods. Secondly, it showed that students are introspectively unaware of their own learning capabilities. Not only did they largely overestimate the value of these approaches, and weren’t able to rank the relative efficiency of each approach. 

Research shows that retrieval doesn’t merely access the stored knowledge in one’s mind and the process of reconstructing information itself enhances learning. Especially when considering the evolving definition of knowledge that values interpretation and analysis, an approach that focuses on practice and application of encoded information seems intuitively synergistic. This new perspective on the human mind opens an exciting set of possibilities regarding the design of new retrieval-based educational activities.