Hierarchical drift diffusion modeling uncovers multisensory benefit in numerosity discrimination tasks.

Area of Science:

• Cognitive Psychology
• Neuroscience
• Decision Making Research

Background:

• Decision-making studies often show a speed-accuracy tradeoff, where improving one metric worsens the other.
• This tradeoff can obscure the benefits of multisensory information when performance is analyzed separately.
• Drift diffusion models (DDMs) can analyze both speed and accuracy simultaneously but typically require large datasets.

Purpose of the Study:

• To investigate multisensory advantages in auditory-visual numerosity discrimination using hierarchical drift diffusion models (HDDMs).
• To demonstrate that HDDMs can reliably estimate parameters with modest sample sizes, overcoming DDM limitations.

Main Methods:

• Utilized hierarchical drift diffusion models (HDDMs) for parameter estimation.
• Applied HDDMs to data from auditory-visual numerosity discrimination tasks.
• Employed hierarchical Bayesian estimation to address sample size constraints.

Main Results:

• Revealed a significant multisensory advantage in auditory-visual numerosity discrimination.
• Demonstrated reliable parameter estimation using HDDMs with a modestly sized dataset.
• Showcased the efficacy of HDDMs in overcoming the large sample size requirements of traditional DDMs.

Conclusions:

• Hierarchical drift diffusion models (HDDMs) are effective for analyzing decision-making tasks involving speed-accuracy tradeoffs.
• HDDMs successfully identified a multisensory advantage in numerosity discrimination.
• This approach makes complex cognitive modeling more accessible by reducing sample size dependency.