Possible influences on the reliability of methods used in the demonstration of cortical predictive coding
MetadataShow full item record
While not being empirically confirmed, cortical predictive coding is a theory of human cognition that has already been used as an explanation for multiple previously unexplained phenomena. In order to empirically demonstrate the physiological implications of predictive coding we would need reliable and corresponding data. The current literature study examines which methodologies could be used for the acquiring of corresponding data and whether acquired data could be interpreted in a different way. This study does this by: 1. describing the physiological implications of cortical predictive coding; 2. examining the strengths and weaknesses of methods that are currently being used in the demonstration of cortical predictive coding and; 3. examining broader influences on data without a one-to-one correlation to a specific measuring tool. The results suggest that there exist multiple difficulties when using any of the discussed methods of research, as some lack spatial resolution (functional magnetic resonance imaging, mismatch negativity) or temporal resolution (single-cell recordings), are sparsely available (invasive electroencephalography), are in development (stimulus specific adaptation, laminar functional magnetic resonance imaging), or are difficult to interpret because of difficulties in substantiating their relevance (mismatch negativity, stimulus specific adaptation). Data seems to be dependent on the used methodology as well as attentional levels, arousal, neuromodulators and the use of animal subjects. Further research should take the methodological weaknesses into account, preferably using multiple methodologies simultaneously to cover for the weaknesses of each individual method. Further research is also encouraged to look into multiple predictive coding properties within one study, thus making it possible to discriminate between the multiple properties.