The κ statistic in rehabilitation research: An examination

摘要: The number and sophistication of statistical procedures reported in medical rehabilitation research is increasing. Application the principles methods associated with evidence-based practice has contributed to need for practitioners understand quantitative published articles. Outcomes measurement determination reliability are areas that have experienced rapid change during past decade. In this study, distinctions between agreement examined. Information presented on analytical approaches addressing focus application K statistic. following assumptions discussed: (1) should be used data measured a categorical scale, (2) patients or objects categorized independent, (3) observers raters must make their decisions judgments independently. Several issues related using studies described, including use weighted K, reporting effect bias prevalence sample size power requirements K. statistic useful assessing among raters, it being more frequently research. Correct interpretation depends meeting required accurate reporting.