Caution: Model mimicry! How to eliminate the risk of the wrong model providing the right data:

Active areas of the brain generate local electromagnetic fields that can be detected on the surface of the skull. For example, in electroencephalography (EEG), an electrode cap worn on the head of patients or study participants is used to measure the electrical component of the fields generated by neurons. This method allows very fast processes in the brain to be measured.

Non-invasive methods therefore play an important role in research into many neuroscientific and neurological research questions. However, scientists at the Ernst Strüngmann Institute have now shown in a study recently published in Nature Communications that it is also essential to classify the data obtained correctly. Otherwise, there is a risk of drawing hasty conclusions or even misinterpreting results.

This is because many different underlying neural effects can lead to identical EEG patterns. And although a given model may explain the observed data well at first glance, there is a risk that it may be based on improbable assumptions about the underlying neural processes.

To prevent this “model mimicry” effect from leading to false conclusions, the researchers point out how important it is to acknowledge the limitations of noninvasive neural recordings such as EEG and use additional methods or data from other sources. This can include, additional experiments, but also data from other approaches such as fMRI and single neuron recordings.

Click here for the original paper