No abstract

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.

The habituation paradigm has been applied to study the development of memory, perception, and other cognitive processes in preverbal infants, making it one of the most prominent experimental paradigms in infant research. However, there are many features of the process of habituation that remain elusive, which results in uncertainty about the best research practices. This article first discusses current practices in habituation research (e.g., the use of habituation criteria) in relation to modelling the process of habituation, revealing several issues that impede progress in the field. To overcome these challenges, we propose to move towards a modelling framework to study critical features of the habituation process. To facilitate this transition, alternative experimental designs are proposed. The article encourages clearer thinking about the process of habituation, such that the theory, design, and analysis are all in line with each other. The article concludes with concrete recommendations to improve current practices in infant habituation research.

Methodological variations and inconsistency in reporting practices pose considerable challenges to the interpretation and generalizability of outcomes derived from the habituation paradigm - one of the most prominent methods for studying infant cognition. In a systematic review, we map out experimental design choices in habituation study samples aged 0-18 months using looking time measures. 2,853 records published in peer-reviewed journals between 2000-2019 were extracted from PsycInfo and Web of Science. 785 (27.5%) papers were identified after screening (Fleiss kappa = .60, 95%, CI[.40 - .80], 6 blind raters). We adopt a collaborative, multi-lab approach for crowd-sourced data collection involving raters from the developmental research community. In a meta-analysis, we assess the impact of habituation detection criteria on the novelty effect size, moderated by age. Our results will inform a detailed evaluation of experimental designs and a set of specific guidelines to improve research and reporting practices in infant habituation research.