Patient MTurk Ratings by Phrase.

Pilot study comparing speech language pathologist (SLP) gold standard ratings to MTurk user ratings stated as means (standard error of the mean). The different phrases are represented by sentence codes KC, TT, SD, PP, SS, ZC. Blank cells indicate that the SLP did not ask the patient to speak those phrases, and therefore there was no recording of that phrase for MTurk users to rate. When an SLP provided a range of scores for a patient, such as for Patient 1, we took the mean of the range as the gold standard (e.g. a score of “2–3” was converted to 2.5).