This project contains materials, data, analysis scripts, and papers/presentations for an extremely large-scale human simulation paradigm experiment investigating how the semantic information that is carried by a word's linguistic contexts is modulated by the kind of nonlinguistic context that word is used in. These experiments were designed and run by Aaron Steven White using Ibex. The experiments were deployed natively on Mechanical Turk using a custom modification of Ibex.
Portions of this work were presented at the the 2014 Congress of the International Association for the Study of Child Language and the 2016 CUNY Conference on Human Sentence Processing. A manuscript describing all the experiments found here is now available on Aaron Steven White's website and in the papers/ directory.
This directory contains all the materials needed to run the experiment. This includes item generation and Ibex templating scripts (create_items.py). The experiments were run natively on Mechanical Turk by extracting the relevant Ibex javascript components. (Thanks to Pranav Anand who helped with this.)
This directory contains the raw data file pulled from Mechanical Turk. results.preprocessed was generated from the raw results files using preprocess.py in the analysis/ directory.
This directory contains the analysis script for the manuscript found in papers/ as well as a preprocessing/download script for use on the raw Ibex data.
This directory contains a paper and posters based on the data.