Crowd-sourced Experiments on Pronominalization in the Free vs Constrained Completion Paradigm
This dataset contains the data and scripts of three experiments on referring expression choice (pronominalization). The experiments were conducted via crowd-sourcing; we aimed to observe whether referent predictability (manipulated by the choice of IC or ToP verbs) affects pronominalization. The experiments use different paradigms.
Steps to reproduce
The data consists of prompts and completions collected via crowd-sourcing. The descriptions of the materials and procedure are included in the linked publication.