Modelling event-related skin conductance responses.

Area of Science:

• Psychophysiology
• Signal Processing
• Mathematical Modeling

Background:

• Psychophysiological signal analysis tools often rely on unstated assumptions.
• Skin conductance responses (SCRs) are a key psychophysiological measure.
• A formal mathematical framework is needed to clarify analysis assumptions.

Purpose of the Study:

• To empirically test the assumption that SCRs can be modeled as the output of a linear time-invariant (LTI) filter.
• To develop and validate a mathematical framework for SCR analysis.
• To investigate the nature of baseline variance and nonlinear interactions in SCRs.

Main Methods:

• Developing a mathematical framework for SCR analysis based on LTI filter principles.
• Empirically testing the LTI model assumptions using SCR data.
• Quantifying variance attributable to evoked responses versus baseline.
• Investigating nonlinear interactions between temporally overlapping SCRs.

Main Results:

• A significant portion of SCR variance is explained by individual-specific response functions.
• Baseline variance in SCRs is greater than variance unexplained by the LTI model.
• No evidence of nonlinear interactions among temporally overlapping evoked SCRs was found.
• A canonical response function was developed and shown to be applicable across different recording sites.

Conclusions:

• The linear time-invariant filter model provides a robust framework for analyzing SCRs.
• Individual response functions are key to understanding SCR variability.
• The findings support model-based analysis of SCRs using established signal processing techniques.