Common Reliability Errors in Kinesiology Research
Without quantitative measurements, no scientific field can make significant progress and, without demonstrated reliability, no measurement can be trusted. Reliability, known also as the dependability of scores, relative freedom from error, stability, internal consistency, reproducibility, agreement, association, sensitivity, precision and objectivity, etc.,is perhaps the most important quality of any measurement. This is because, without it, there would be no validity,another very important quality of a measurement. Suppose an examinee took a test once and then was able to go back in time as if. The test had never been taken. Then suppose the same test was administered again. We would expect the scores on the two tests to be quite similar. If not, the test would be unreliable and the information we learned about the examinee will be useless. If a test is administered at the end of a unit of instruction, we assume the scores will represent a stable level of achievement. In other words, if the test were administered again the next day, the test scores would be similar to those obtained on the previous day.
Weimo Zhu
University of Illinois at Urbana-Champaign, IL, USA Guangzhou Sport University
国际会议
The 2008 International Sports Biomechanics Symposium(2008国际运动生物力学研讨会)
广州
英文
34-37
2008-07-30（万方平台首次上网日期，不代表论文的发表时间）