Clozapp - McMaster Experts

Abstract This paper introduces a freely available and easy to use Java application for the collection and recording of Cloze probability ratings. Clozapp presents participants with text fragments of the researchers’ choice and collects guesses regarding upcoming words. It can also collect basic demographic information about participants. Available modes of data collection include elicitation of responses to a limited number of omitted words in a text or to all words in a text. Clozeapp can be customized to present instructions and experimental stimuli in any given language and to collect multiple types of demographic data. This paper presents the application by detailing the states and actions available, as well as descriptions of how to customize the app to fit different experimental needs including possible input and output details. The application manual is provided. As a proof of concept, we used Clozapp to conduct a replication study of two existing collections of Cloze probability norms. The Clozapp norms showed strong reliable correlations (r > 0.7) with both existing data sets, suggesting a high convergence between modes of data collection. The application provides an efficient and customizable way of collecting predictability norms for language research.

Clozapp Journal Articles