Lecture 6 - The Event-Related Potential (ERP) Method using EEG Flashcards
What is an Event-Related Potential (ERP) in regards to EEG?
An Event-Related Potential (ERP) is defined as the signal picked up by EEG electrodes that is related to a specific cognitive event.
In order to define/measure an ERP what assumptions need to be made?
The event of interest (the cognitive event) is defined in time.
The event consistently evokes the signal of interest.
The timing of the signal is consistent.
The signal and the noise are uncorrelated.
The noise is random with a mean of zero.
How are ERPs defined/measured?
Hint: think of what a spectrogram of one trial looks like compared to an overlay of many spectrogram of replicated trials.
ERPs are defined by averaging amplitude data of the signals/spectrograms of many replicated trials. When we do this the noise is averaged out and what we are left with is a clearer estimate of an ERP.
How are ERPs described?
ERPs are described in terms of polarity and order of the peaks and trophs.
The peaks and trophs along with their polarity and order are defined as the “components” of an ERP.
So, for a given ERP there will be a set of P1,P2,P3, and N1, N2, N3 representing the components of the ERP.
What is the dependent variable for ERP analysis?
The dependent variable in ERP analysis is amplitude of the ERP.
Note: the way in which amplitude is measured varies between experimenters and how the amplitude is measured will influence the type of results obtained.
What are some different ways in which amplitudes can be measured in ERP analysis?
Baseline-to-peak (most common).
Peak-to-peak.
Area under the curve.
Latency (which refers to the time of onset of the ERP component).
Describe one way in which ERPs have been used to study a cognitive process.
Error-Related Negativity (ERN) refers to a cognitive process that reflects a change in brain activity in response to making a mistake/error. The ERN is a component on an ERP that can be measured.
Gehring et al (1993) used ERPs to study the cognitive process of ERN to understand whether ERN is only reflective of recognition of a mistake/error or if it was also associated with further changes in behaviour in response to this recognition.
They found that after recognising that they had made a mistake, participants engaged with the subsequent question with less force, with more time, and with higher accuracy. This provided evidence that ERN is not just about recognition but also about changing behaviour in response to this recognition of error.
How useful are ERPs in measuring cognitive processes?
ERPs are likely to represent many different cognitive processes occurring at the same time and therefore an ERP cannot be considered to be representative of single cognitive task. Furthermore, ERPs may lend themselves easily to reverse inference. Lastly,ERP analysis varies amongst researches and therefor one needs to be careful in how and why inferences
How useful are ERPs in measuring cognitive processes?
ERPs are likely to represent many different cognitive processes occurring at the same time and therefore an ERP cannot be considered to be representative of single cognitive task. Furthermore, ERPs may lend themselves easily to reverse inference. Lastly,ERP analysis varies amongst researches and therefor one needs to be careful in how and why inferences are being made based on an ERP.
Can ERPs be used to describe which cognitive processes are occurring and/or predict what a EEG trace will look like based on what cognitive task is being performed?
Yes. There are certain ERPs that are well defined and have been used to develop computer programs that can, based on an EEG trace, determine what cognitive task is being performed, and also predict, based on what cognitive task is being performed, what the EEG trace will look like.
What assumptions are made when generating an ERP trace?
There are 5 we talked about
- The event of interest is defined in time
- The event consistently evokes the signal of interest
- The timing of the signal is consistent.
4, The signal and the noise are uncorrelated - The noise is random with a mean of zero.
How is an ERP defined/obtained?
Amplitude data of many trials of the same person doing the same task are averaged.
How are ERP described?
By polarity and order.
When defining an ERP what is the most common form of measuring it?
Using the baseline-to-peak as a measurement of amplitude.
In 1993 Gehring et al. used ERPs to define the ERROR-RELATED NEGATIVITY (ERN).
What did they find out about ERN and how it was related to mistake recognition and compensation?
They found that a higher amplitude of ERP was associated with lower response force, higher probability of getting the next round right, the slower the subsequent response.
