Cognitive strategies revealed by clustering eye movement transitions

Autor(es)

Kucharsky, Simon ^[1] ; Visser, Ingmar ^[1] ; Trutescu, Gabriela-Olivia ^[2] ; Laurence, Paulo G. ^[3] ; Zaharieva, Martina ^[1] ; Raijmakers, Maartje E. J. ^{[1, 4]}

Afiliação do(s) autor(es)

^[1] Univ Amsterdam, Amsterdam - Netherlands

^[2] Leiden Univ, Leiden - Netherlands

^[3] Univ Prebiteriana Mackenzie, Sao Paulo - Brazil

^[4] Vrije Univ, Amsterdam - Netherlands

• Tipo de documento: Artigo Científico
• Fonte: JOURNAL OF EYE MOVEMENT RESEARCH; v. 13, n. 1 2020.
• Citações Web of Science: 0

Resumo

In cognitive tasks, solvers can adopt different strategies to process information which may lead to different response behavior. These strategies might elicit different eye movement patterns which can thus provide substantial information about the strategy a person uses. However, these strategies are usually hidden and need to be inferred from the data. After an overview of existing techniques which use eye movement data for the identification of latent cognitive strategies, we present a relatively easy to apply unsupervised method to cluster eye movement recordings to detect groups of different solution processes that are applied in solving the task. We test the method's performance using simulations and demonstrate its use on two examples of empirical data. Our analyses are in line with presence of different solving strategies in a Mastermind game, and suggest new insights to strategic patterns in solving Progressive matrices tasks. (AU)

• Processo FAPESP: 18/09654-7 - Relacionando estratégias cognitivo-visuais e funções executivas: exemplos com eye-tracking em tarefas de raciocínio indutivo e leitura
• Beneficiário: Paulo Guirro Laurence
• Modalidade de apoio: Bolsas no Brasil - Doutorado