The Kappa coefficient is used for correlation testing of categorical data, while the Pearson correlation coefficient is used for correlation testing of quantitative data. The general Kappa coefficient can be used to measure classification accuracy or categorical data such as voting.

    Background description:

    The purpose of Kappa consistency testing is to evaluate the paired consistency between different indicators. By calculating Kappa values and related statistical indicators, we can understand the consistency between each pair of indicators. Kappa value and p-value are important indicators for measuring pairing consistency. The closer the Kappa value is to 1, the higher the pairing consistency, while the smaller the p-value, the more significant the pairing consistency. These results can help us determine the most suitable evaluation method among different indicators.

    The analysis results are as follows:

    According to Kappa analysis,

    Based on variables A1 and A2, it can be seen that the significance P-value is 0.014, showing significance at the level. Rejecting the original hypothesis indicates extremely low consistency between the two variables. Meanwhile, the value of Kappa coefficient is 0.070.

    Based on variables A1 and A3, it can be seen that the significance P-value is 0.000, showing significance at the level. Rejecting the original hypothesis indicates moderate consistency between the two variables. Meanwhile, the value of Kappa coefficient is 0.492.

    Based on variables A1 and A4, it can be seen that the significance P-value is 0.000, showing significance at the level. Rejecting the original hypothesis indicates low consistency between the two variables. Meanwhile, the value of Kappa coefficient is 0.356.

    Based on variables A1 and A5, it can be seen that the significance P-value is 0.000, showing significance at the level. Rejecting the original hypothesis indicates low consistency between the two variables. Meanwhile, the value of Kappa coefficient is 0.180.

    Based on variables A2 and A3, it can be seen that the significance P-value is 0.001, showing significance at the level. Rejecting the original hypothesis indicates low consistency between the two variables. Meanwhile, the value of Kappa coefficient is 0.125.

    Based on variables A2 and A4, it can be seen that the significance P-value is 0.698, which does not show significance at the level and cannot reject the original hypothesis, indicating extremely low consistency between the two variables. Meanwhile, the value of Kappa coefficient is -0.002.

    Based on variables A2 and A5, it can be seen that the significance P-value is 0.004, showing significance at the level. Rejecting the original hypothesis indicates extremely low consistency between the two variables. Meanwhile, the value of Kappa coefficient is 0.096.

    Based on variables A3 and A4, it can be seen that the significance P-value is 0.000, showing significance at the horizontal level. Rejecting the original hypothesis indicates low consistency between the two variables. Meanwhile, the value of Kappa coefficient is 0.392.

    Based on variables A3 and A5, it can be seen that the significance P-value is 0.000, showing significance at the level. Rejecting the original hypothesis indicates low consistency between the two variables. Meanwhile, the value of Kappa coefficient is 0.262.

    Based on variables A4 and A5, it can be seen that the significance P-value is 0.001, showing significance at the level. Rejecting the original hypothesis indicates low consistency between the two variables. Meanwhile, the value of Kappa coefficient is 0.130.

    From the results, it can be seen that the pairing consistency between A1 and A2 is low, while the pairing consistency between A1 and A3, A1 and A5 is high. In addition, the pairing consistency between A1 and A4 was not significant. These results can help us evaluate the reliability and consistency of different indicators, thereby providing more accurate data analysis and interpretation.

    Reference:

    [1]唐万, 胡俊, 张晖,等. Kappa 系数:一种衡量评估者间一致性的常用方法(英文)[J]. 上海精神医学(Shanghai Archives of Psychiatry), 2015(1):62-67.

    [2] 夏邦世, 吴金华. Kappa一致性检验在检验医学研究中的应用[J]. 中华检验医学杂志, 2006, 29(1):83-84.     [3]Maritz. J.S. (1981) Distribution-Free Statistical Methods, Chapman & Hall. ISBN 0-412-15940-6. (page 217)

关注微信公众号发送【示例数据】获取SPSSMAX练习示例数据。