2020 Thorson ControllingUncontrolled
The article investigates whether incidental characteristics and behaviors of experimenters influence participants' peripheral autonomic nervous system (ANS) activity during baseline recordings and task‑induced reactivity. Across nine independent datasets (1,341 participants, 160 experimenters) the authors apply multilevel modeling that treats experimenter identity as a random effect and experimenter status (junior vs. senior), gender, and race as fixed effects. Physiological indices include pre‑ejection period (PEP), inter‑beat interval (IBI), respiratory sinus arrhythmia (RSA), and root‑mean‑square of successive differences (RMSSD). Random‑effects analyses reveal virtually no significant between‑experimenter variance in baseline or reactivity measures, with a single marginally significant RMSSD variance in one study. Fixed‑effects tests show no reliable impact of experimenter status, gender, or race on any ANS outcome, aside from isolated interactions (e.g., male participants showing higher baseline RSA with male experimenters, and same‑race dyads showing lower baseline RSA). The authors argue that the lack of systematic effects may stem from limited variability in experimenter characteristics, low physiological reactivity in some paradigms, and brief experimenter‑participant contact. They provide analytic syntax, recommend documenting experimenter attributes, using multiple experimenters, counterbalancing, and incorporating self‑report items to detect subtle influences, and suggest that future work should examine more visible experimenter cues, larger variance in participant susceptibility, and other physiological systems.
Topics
Analytic Framework for Detecting Experimenter Effects in Psychophysiology
Presentation of multilevel modeling strategies—random intercepts for experimenter identity and fixed effects for status, gender, and race—to quantify how much variance in ANS measures can be attributed to experimenters.