Javascript must be enabled for the correct page display

Optimize learning with reaction time based spacing

Thiel, W.G.E. van (2010) Optimize learning with reaction time based spacing. Master's Thesis / Essay, Human-Machine Communication.

AI-MHMC-2010-w.g.e.vanThiel.pdf - Published Version

Download (605kB) | Preview


Can we optimize learning efficiency by modifying the order of items in a learning session? By taking into account well-known memory phenomena, we can improve learning. However, in practice, learning methods that take into account memory effects such as primacy, recency and spacing are not often used. Especially the characteristics of the spacing effect, which refers to enhanced learning when trials are spaced over time instead of massed in a short time, are only rarely applied. This study proposes an adaptive cognitive model that takes these effects into account and is easily applicable in practice. This model lets people learn facts effectively and is a refined version of models from previous research by Van Woudenberg (2008) and Pavlik and Anderson (2008). Just as in these studies, the new model keeps a representation of the strengths of items in memory. As in the study of Van Woudenberg (2008), the memory strengths are based on response times of the user, but now based on a more direct and continuous measure. On the basis of the strength representations and a forecast of the development of these strengths, the model optimizes the word order. This model is compared to a standard teaching schedule in an experiment done with three Havo/VWO classes. In a learning session of fifteen minutes, students studied twenty Dutch-French word pairs. The next day, performance of these word pairs was tested. The training data of the experiment showed that the model’s prediction of response times, e.g. its representation of strength of items in memory, improved as repetition increased. Analyses of the test data showed that participants in the adaptive condition scored on average 1.1 point higher on a scale of 1 to 10 than the control condition. Although more refinements are still possible, this work confirmed that spacing through adaptation based on reaction times yields an effective learning method.

Item Type: Thesis (Master's Thesis / Essay)
Supervisor name: Rijn, H. van
Degree programme: Human-Machine Communication
Thesis type: Master's Thesis / Essay
Language: English
Date Deposited: 15 Feb 2018 07:31
Last Modified: 02 May 2019 12:45

Actions (login required)

View Item View Item