According to ERP article from Wikipedia, it is mentioned that "To see the brain's response to a stimulus the experimenter must conduct many trials and average the results together, causing random brain activity to be averaged out and the relevant waveform to remain, called the ERP". Does it mean we need to show the visual stimulus to the subjects a number of times, create epochs of the individual trials and then the averaging ?
Yes, in order to average out the noise in the EEG data, you need multiple epochs time locked to the stimuli of interest. For instance, in an experiment examining the brains response to semantic anomalies (N400), around 30-40 trials per condition is usually required. In other experimental paradigms such as e.g. MMN, more trials might be needed, due to lower effect size and therefore lower signal-to-noise ratio for the ERP component of interest.
In my experiment, I have display of two faces on left and right side of the screen. The subject needs to press the button accordingly to send a response for which image is being repeated(Go paradigm). Similarly if the desired stimulus is not shown (No-go paradigm).
The multiple trials for ERP in this case refers to the repetition of the condition but using different images right ?
For example, Taking Image(A) as the target stimuli:
For Go paradigm:
In the 1st trial image is A:Left-B:Right
In the 2nd trial, image is A:Left-C:Right
In the 3rd trial, image is D:Left-A:Right
and the subject needs to choose A accordingly.
For No-go paradigm:
In 4th trial, the image is E:Left-F:Right
In 5th trial, the image is G:Left-H:Right and so on.
and the subject needs to avoid choosing any button.
So my understanding is the 1st, 2nd, 3rd, 4th and 5th trials refers to as multiple trials right ? Thomas Hörberg . Thank you
Yes, but in this case 1st, 2nd, and 3rd trials are single trial epochs of the go-condition, and 4th and 5th trials are single trial epoch of the no-go-condition.
- Badges
- Science topic