Logistic analysis of choice data: A primer.

Area of Science:

• Decision Neuroscience
• Behavioral Economics

Background:

• Logistic regressions originated in economics for modeling individual choice behavior.
• These models have become increasingly important in decision neuroscience research.

Purpose of the Study:

• To describe and discuss various logistic models used in decision neuroscience.
• To emphasize the assumptions and interpretations of these logistic models.
• To highlight their application in quantifying behavioral traits and neuronal influences on choices.

Main Methods:

• Review and discussion of different logistic regression models.
• Application of logistic models to quantify subjective value, choice accuracy, risk attitudes, and biases.
• Extension to complex scenarios including bundle choices, nonlinear value functions, and multiple options.
• Utilizing logistic models to assess the explanatory power of neuronal activity on choices.

Main Results:

• Logistic models can quantify diverse behavioral traits, including subjective value, risk attitudes, and choice biases.
• Complex logistic models accommodate choices involving bundles, nonlinear value functions, and multiple options.
• Logistic models provide a quantifiable link between neuronal activity and decision-making.
• These models offer a viable alternative to receiver operating characteristic (ROC) analyses for evaluating neuronal influence.

Conclusions:

• Logistic regression models are versatile tools in decision neuroscience for understanding choice behavior.
• They offer a robust framework for quantifying behavioral economics principles and neuronal correlates of choice.
• Logistic models present a powerful alternative to traditional methods like ROC analysis for dissecting decision processes.