[Some exact results for random walk models with applications].

Area of Science:

• Decision Sciences
• Mathematical Psychology
• Statistics

Background:

• Wald's approximation is commonly used for random walk models but neglects excess over absorbing barriers.
• Existing models like the Sequential Probability Ratio Test (SPRT) and Simple Stopping Rule (SSR) have limitations.

Purpose of the Study:

• To present a random walk model that provides exact predictions without relying on Wald's approximation.
• To evaluate the model's performance against experimental data.
• To clarify the model's relationship with SPRT and SSR models.

Main Methods:

• Development of a random walk model analyzed without Wald's approximation.
• Derivation of exact predictions for absorption probabilities and mean conditional absorption times.
• Fitting the model to experimental data from Green et al. (1983).

Main Results:

• The proposed model yields exact predictions for absorption probabilities and mean conditional absorption times.
• The model demonstrated a satisfactory fit to the experimental data.
• The model was shown to belong to the SPRT or SSR classes, or both, depending on parameter values.

Conclusions:

• The new random walk model offers a more accurate alternative to Wald's approximation.
• Exact results are theoretically significant for evaluating approximations and understanding random walk models.
• The model provides a valuable tool for analyzing decision-making processes and experimental data.