That's a semantic quibble which isn't really relevant to the point that Harrison is making in the rest of the quoted paragraph.
But yes, you can have misplaced faith in e.g. null hypothesis significance testing and other methods of data collection and analysis common in some of the social sciences.
Using faith and belief interchangeably seems disingenuous and ignores the religious context that is part of faith. Faith implies a lack of compromise, a concreteness, and unwillingness to change. Using that term to describe evidence based science, and the belief that evidence ought to be presented before a belief is held is an attempt to reduce science to dogma.
What you describe is more like fideism.