Text this: How the different explanation classes impact trust calibration: The case of clinical decision support systems