LaStayo and Wheeler1 are to be commended for their investigation on goniometric measurements at the wrist. Clearly, their work provides evidence for using a specific approach—dorsal and volar placements of the goniometer—for measuring wrist flexion and extension. Such work is the cornerstone of evidence-based practice. After reading their article, however, I have several questions concerning the analysis. Specifically, the questions focus on the types of intraclass correlation coefficient (ICC) chosen by the authors and their interpretation of the work of Shrout and Fleiss.2 My choices would have been to use the type (1,1) ICC for the intertherapist design and the type (2,1) for the intratherapist design (test-retest design), as the variation can be due to either the therapist or the patient, or more likely the combination of the two.… Paul C LaStayo, PT, CHT Instructor, Department of Physical Therapy, College of Health Related Professions, University of Florida, JHMHC Box 100154, Gainesville, FL 32610 Donna L Wheeler, PhD Research Assistant, Department of Orthopaedics, University of Florida