ERPs in single trials are in my opinion rather difficult to obtain from single trials. The strength of ERPs lies in their repetitive recording and the subsequent averaging over trials in order to obtain a clean signal that does not contain any information irrelevant to the task.
This can be seen when considering the earliest studies on ERPs, as done for instance by Kornhuber and Deecke 1965. Even their recording of the readiness potential required numerous repetitions and motor potentials are usually quite high in their amplitude, meaning fairly easy to extract.
Single trial ERPs will require an extraordinarily clean signal, with no noise at all. Given that such circumstance are hard to control in single trials, it is probably way easier to simply record several trials instead.
However, this is only an answer to ERPs in general and the exact nature of your inquiry in terms of what you want to measure is key in order to answer your question specific to what you want to know.
Hi Prof. Aljoscha Thomschewski, Aljoscha Thomschewski
Thank you for your kind reply.
I would like to try to build a mixed-effect model for ERP amplitude through the R software, which would need to take into account the random effects of SUBJECT and ITEM, so I think it is necessary to extract the amplitude of a single trial.
However, at present, in the field of psycholinguistic research, there is little literature on single-trial analysis, and there is no mention of how to extract the single-trial waveforms. If the method is the same as traditional repetitive and averaging ERP preprocessing procedure, although theoretically I think it should be the same, it is really huge work.
So I would like to ask if any teachers have done this who can share the single-trial analysis method. I am also not sure if my understanding is correct as there are limited resources available.