O'Connor P, Lorey R E
J Dent Educ. 1978 Apr;42(4):174-9.
Research shows that use of rating scales with precisely defined scale points improves interrater agreement. Instead of words, scale points may be defined by products for which ratings have been established by consensual agreement. These products serve as standards for raters. In two studies, raters assessed student products using conventional rating scales and identical scales together with comparison stimuli (photographic exemplars) of consensually established values of 3 (adequate) for each criterion. Results of the first study (but not the second) clearly confirmed the contribution of exemplars to interrater agreement. Findings suggested that single exemplars significantly enhance agreement for criteria on which the ideal is an end point on a continuum. Multiple exemplars may be necessary for other criteria. Comparison stimuli offered promise for improving interrater agreement, but further work is necessary to adapt their use for complex criteria that frequently define the quality of dental preparations.
研究表明，使用具有精确定义量表点的评定量表可提高评分者间的一致性。量表点可以由已通过共识确定评级的产品来定义，而不是用文字。这些产品作为评分者的标准。在两项研究中，评分者使用传统评定量表和相同量表，以及每个标准的共识确定值为3（足够）的比较刺激物（照片范例）来评估学生作品。第一项研究（而非第二项）的结果清楚地证实了范例对评分者间一致性的贡献。研究结果表明，单个范例可显著提高理想情况是连续体终点的标准的一致性。对于其他标准，可能需要多个范例。比较刺激物有望提高评分者间的一致性，但需要进一步开展工作，以使其适用于经常定义牙科预备质量的复杂标准。
