Enhancing simulations with intra-subject variability for improved psychophysical assessments

Area of Science:

• Psychology
• Psychophysics
• Statistics

Background:

• Psychometric properties like reliability are crucial for perceptual assessments.
• Reliability is influenced by method, inter-subject, and intra-subject variability.
• Intra-subject variability is often neglected due to quantification challenges.

Purpose of the Study:

• To develop a novel approach for estimating intra-subject variability in psychometric functions.
• To account for intra-subject variability in simulating experiments and predicting reliability.
• To guide efforts in optimizing assessment procedures versus addressing confounding factors.

Main Methods:

• Combined computer simulations with behavioral data to model intra-subject variability.
• Applied the approach to proprioceptive difference threshold assessments.
• Utilized two measurements per subject to estimate intra-subject variability.

Main Results:

• Simulations neglecting intra-subject variability overestimated reliability (r=0.768 vs. actual r=0.212).
• The new approach accurately predicted reliability (r=0.207) by including intra-subject variability.
• Identified that reducing intra-subject variability is key, not just optimizing sampling procedures.

Conclusions:

• The proposed method provides realistic reliability estimates by accounting for intra-subject variability.
• It enables prediction of reliability for larger cohorts and retests without further experiments.
• This tool is valuable for time-constrained populations, particularly in clinical settings.