Text this: Are evaluations in simulated medical encounters reliable among rater types? A comparison between standardized patient and outside observer ratings of OSCEs