SweetPea: A standard language for factorial experimental design

Area of Science:

• Empirical Research Methodology
• Computational Science
• Psychological Research

Background:

• Reproducible empirical research relies heavily on robust experimental design.
• Increasing complexity in experimental designs poses challenges for researchers in avoiding confounds.
• Existing methods may not adequately address the need for unbiased sampling of experimental sequences.

Purpose of the Study:

• To introduce SweetPea, an open-source declarative Python language for experimental design.
• To demonstrate SweetPea's utility in designing psychological experiments.
• To explore SweetPea's potential applications in other scientific domains.

Main Methods:

• Utilizing a declarative language (Python) to specify experimental factors and constraints.
• Leveraging computer science advancements for unbiased sampling of experiment sequences.
• Applying SweetPea to the practical design of psychological experiments.

Main Results:

• SweetPea enables researchers to define complex experimental parameters declaratively.
• The system facilitates unbiased sampling, crucial for reproducible results.
• Demonstrated successful application in psychological experiment design.

Conclusions:

• SweetPea offers a powerful tool for improving the reproducibility of empirical research.
• The language's flexibility supports diverse experimental needs.
• Potential for expansion into neuroscience and machine learning research designs.