The Magic Curiosity Arousing Tricks (MagicCATs) database in Italian younger and middle-aged adults: Descriptive statistics and rule-based machine learning

TL;DR

This study validates a database of magic-trick videos designed to induce curiosity and uses machine learning to understand how different emotions are linked.

Contribution

The study validates the MagicCATs database for Italian adults and applies rule-based machine learning to analyze epistemic emotions.

Findings

The MagicCATs database was validated for use in Italian younger and middle-aged adults.

Association rule learning revealed co-occurrences between different epistemic emotions.

Descriptive statistics and machine learning aid in stimulus selection for psychological experiments.

Abstract

Epistemic emotions, and in particular curiosity, seem to enhance memory for both the specific information that stimulates the individual’s curiosity and information presented in close temporal proximity. Most studies on memory and curiosity have adopted trivia questions to elicit curiosity. However, the amount and range of interest that trivia questions elicit are unclear, and there is no established, universal trivia item pool guaranteed to elicit comparable levels of curiosity across individuals of all ages. Thus, one substantial challenge when studying curiosity is systematically inducing it in controlled experimental settings. Recently, an innovative database called Magic Curiosity Arousing Tricks (MagicCATs) has been published. This database includes 166 short magic-trick video clips that adopt different materials and is designed to induce curiosity, surprise, and interest. Here, we aimed to validate this dataset in the Italian population by reporting the basic characteristics and the norms of the magic-trick video clips in younger and middle-aged adults. We also carried out association rule learning, a rule-based machine learning and data mining method to aid understanding of the co-occurrences between the different epistemic emotions and aid researchers in stimulus selection. Association rules underline relationships or associations between the variables in our datasets and can be used in association with descriptive statistics for stimulus selection in psychological experiments.