Discounting: A practical guide to multilevel analysis of choice data

Area of Science:

• Behavioral Economics
• Decision Science
• Psychology

Background:

• Understanding reward discounting is crucial for explaining choices involving immediate versus delayed gratification.
• Existing models often struggle to simultaneously account for individual and group-level discounting patterns.
• The interplay between reward magnitude and delay significantly influences decision-making processes.

Purpose of the Study:

• To introduce and advocate for a multilevel modeling approach to analyze reward discounting.
• To simultaneously evaluate discounting behavior at both individual and group levels.
• To differentiate the contributions of relative reward magnitude and relative delay to choice behavior.

Main Methods:

• Utilizing multilevel logistic regression to model choice data.
• Examining choices between smaller sooner and larger later rewards.
• Implementing a framework that treats sensitivity to reward magnitude and delay as distinct factors.

Main Results:

• The proposed multilevel logistic regression model effectively captures individual and group discounting.
• Sensitivity to relative reward magnitude and relative delay are identified as significant, separate predictors of choice.
• The methodology provides a robust framework for analyzing complex choice data.

Conclusions:

• Multilevel modeling offers a powerful tool for researchers studying reward discounting.
• Separating the effects of reward magnitude and delay enhances the precision of discounting models.
• This approach facilitates a deeper understanding of decision-making under conditions of delayed gratification.