The participants’ functioning was rated by 2 physical therapists with the Extended ICF Core Set for Stroke. The physical therapists had at least 5 years of experience in neuro- rehabilitation, had completed sev- eral postgraduate courses in the re- habilitation of motor performance, and were well trained in the use of the ICF. The ICF training consisted of workshops, a pilot phase with dis- cussion of cases before initiation of the study, and the supervision of an expert from the ICF Research Branch, Munich, Germany. The rat- ings were based on information from the participants (interview and ob- servation), from interviews with proxies (eg, a spouse, partner, or close relative, who also could have the role of a caregiver), from various health care professionals, and from medical records containing the results of standardized examinations. This approach was chosen because the accuracy and comprehensiveness of raters’ clinical judgments are expected to increase when various sources of information and various perspectives are taken into account. Furthermore, people with neurological conditions often are not optimally situated to judge the severity of their deficits because of the common phenomenon of unawareness.31 Both physical therapists attended the interviews. The role of the interviewer was assigned at random. The 2 physical therapists completed the Extended ICF Core Set for Stroke in- dependently of each other. For each category contained in the Extended ICF Core Set for Stroke, the raters judged their confidence in the ratings on a scale from 0% to 100%. The scale was subdivided into 10% steps.