Because we are interested in estimating individual (and item) differences using models--e.g., using eyetracking growth curves as differences in processing speed or accuracy on speech production tasks by item or child--we should add these to the database.
We will also have to add the underlying raw data, etc, etc.