Companion repository for the 2022 article "Generalizing Syllogistic Reasoning: Extending Syllogisms to General Quantifiers" published in the proceedings of the 44rd Annual Meeting of the Cognitive Science Society.
analysis
: Contains the analysis scripts generating the results and figures from the paper.analysis/comparison.py
: Script for comparing the Ragni2016 dataset (traditional syllogisms) to the Ragni2021 dataset (generalized quantifiers). Generates a plot visualizing both patterns (Figure 1) and calculates the RMSE.analysis/correct_conclusions.py
: Helper file containing a dictionary with the logically correct responses for all 144 syllogisms with the quantifiers All, Some, Some not, No, Most and Most not.analysis/correctness.py
: Plots the correctness by syllogism type (Figure 3 and Figure 4).analysis/genquant_corr_traditional_corr
: Plots the correctness of participants on traditional and generalized syllogisms (Figure 6).analysis/misinterpreted
: Plots the correctness when accounting for misinterpreted meanings of the quantifiers (Figure 5).analysis/pattern_collapsed.py
: Plots the response behavior with focus on the quantifiers (Figure 2).data
: Contains the datasetsdata/Ragni2016.csv
: Ragni2016 dataset from the CCOBRA framework containing responses to traditional syllogisms.data/Ragni2021.csv
: Responses to all 144 syllogisms with the quantifiers All, Some, Some not, No, Most and Most not.
- Python 3
- CCOBRA
- pandas
- numpy
- matplotlib
- seaborn
- scipy
After downloading the repository, navigate to the analysis subfolder:
cd /path/to/repository/analysis
All scripts can be executed without entering additional parameters. The scripts will create the plots in the same folder. To execute the scripts, enter:
$> python [script].py
Brand, D., Mittenbühler, M., & Ragni, M. (2022). Generalizing Syllogistic Reasoning: Extending Syllogisms to General Quantifiers. In proceedings of the 44rd Annual Meeting of the Cognitive Science Society.