One piece of research that measures vitals instead of just asking the users which one they liked more isn't very convincing. It could even be A/B tested by measuring how long people stay on the page or bounce rate above the fold.