Kropp P, Kiewitt A, Göbel H, Vetter P, Gerber W D
Department of Medical Psychology, University of Kiel, Germany.
Appl Psychophysiol Biofeedback. 2000 Mar;25(1):33-41. doi: 10.1023/a:1009533405695.
Reliability parameters of a test indicate the stability (and quality) of the test itself. Reliability coefficients greater than 0.70 suggest an attribute as being sufficiently stable over time to be characterized as a trait. Reliability parameters of contingent negative variation (CNV) amplitudes in 27 healthy individuals were determined using a test-retest design. CNV was recorded at Cz, with an interstimulus interval of 3 s, on 2 separate occasions: initial session and 10 days later. Correlation coefficients between the 2 recording sessions were 0.675 for the total-CNV (tCNV), 0.855 for the early component (iCNV), 0.631 for the late component (lCNV), and 0.420 for the post-imperative negative variation (PINV). Statistical retest parameters for Spearman Brown were 0.806 for tCNV, 0.922 for iCNV, 0.774 for lCNV, and 0.655 for PINV. The iCNV, more than the other parameters, remained stable over the period of 10 days. It is suggested that the described standardized CNV recording procedure ensures reproducible and stable results in healthy subjects.
一项测试的可靠性参数表明了该测试本身的稳定性（以及质量）。大于0.70的可靠性系数表明一种属性在一段时间内足够稳定，可以被表征为一种特质。采用重测设计确定了27名健康个体的伴随负变化（CNV）振幅的可靠性参数。在Cz点记录CNV，刺激间隔为3秒，分两次进行：初次测试和10天后。两次记录之间的相关系数，总CNV（tCNV）为0.675，早期成分（iCNV）为0.855，晚期成分（lCNV）为0.631，指令后负变化（PINV）为0.420。Spearman Brown统计重测参数，tCNV为0.806，iCNV为0.922，lCNV为0.774，PINV为0.655。与其他参数相比，iCNV在10天内保持更稳定。建议所描述的标准化CNV记录程序在健康受试者中可确保可重复和稳定的结果。
