Electrophysiological Endophenotypes and the Error-Related Negativity (ERN) in Autism Spectrum Disorder: A Family Study.

Ann et al. (2017) wired up kids and adults with and without autism. They measured the error-related negativity, a tiny brain wave that pops up right after a mistake. The goal was to see if this wave could serve as a family marker for autism.

The wave looked the same in both groups. No bigger, no smaller, no delayed timing. Because relatives also showed no special pattern, the team concluded the ERN is not a useful autism endophenotype.

Stevens et al. (2018) found a five-item checklist that does work for cheap screening. Their OERA gave 93 % sensitivity and 91 % specificity in kids aged 3–10. The ERN study used expensive EEG yet came up empty; the OERA used simple questions and succeeded.

Villa et al. (2010) already showed the PEP-R has solid reliability for kids under 12. Like the ERN paper, they asked “does this tool hold up?” The PEP-R passed; the ERN did not.

Nadwodny et al. (2025) and Dudley et al. (2019) tell the same story for LENA. One day of recording gives stable toddler data, but the algorithm fails in older kids. Together these papers warn: a fancy metric is only helpful if it survives real-world use.

Before you buy new EEG gear or add a neural marker to reports, pause. Quick rating scales like OERA or play-based tools like PEP-R already give actionable data for less money and time. Save the lab budget for clients who truly need neuro-assessment, not for routine screening.