> Study included 28 subjects

Well beside the other process issues, 28 is a farcically small set to draw any conclusions, even future hypotheses from.

Scale that to 2800 and we can talk.