Visual cognitive strategies are strategies that a test taker cognitively uses when performing a test to try to do it more efficiently. To study these visual strategies, eye-tracking is a good technique, since it tracks the eye movements. Some methods to analyze the strategy used primarily fixations and gazer time to try to understand those strategies, but recently more sophisticated approaches are emerging. Those methods are based on the use of algorithm and machine learning. Diverse tasks had the strategies studied with those more sophisticated methods, but matrix reasoning tests are the ones with more study. In mathematical tasks, no sophisticated methods to analyze the strategy were found. Therefore, the aim of this project is to develop a sophisticated method to analyze cognitive visual strategy in the mathematical multiplication task based on the techniques used in other tasks, specially matrix intelligence tests. This is very relevant because if a better strategy is diagnosed, it can be taught to underperformers in order to achieve a better performance. To reach our goal, two experiment were designed. In the first, we will study the application of these sophisticated approaches in participants that answered a matrix reasoning test two times, with a 6-week interval between each. This is a favorable situation to understand changes in strategy induced by learning. In the second experiment, we will develop and adapt those sophisticated methods to a mathematical multiplication task based on algorithms that track the transition of the eyes between areas of interest and where the participant engages first.
News published in Agência FAPESP Newsletter about the scholarship: