There is always a striking contrast in brain imaging studies between the precision with which all the parameters of the experiments are controlled (timing, stimuli, etc... ) and the lack of knowledge about the subjects actual cognitive state during the experiment (attention level, subtle emotional reactions, motivation, etc ... ). This lack of knowledge result in a large unexplained variability in the brain imaging data that one collects during such studies.
We discuss here the possibility to ask directly the subjects about their subjective experience during the recordings to collect first-person data that could help reduce the unexplained variabiliy of imaging data. Then starting from these (mostly) methodological considerations, we discuss a larger research project called Neurophenomenology that attempts to build a bridge between first-person and third-person data to better understand the roots of conscious experience.