信度分析怎么选择题目？懿说学区，48

人气：318 ℃/2024-04-16 13:18:00

Yishuo School District (48) SPSS Statistical Analysis (58) Re-test Reliability Analysis

“分享兴趣，传播快乐，增长见闻，留下美好！大家好，这里是小编。欢迎大家继续访问学苑内容，我们将竭诚为您带来更多更好的内容分享。

"Share interests, spread happiness, increase knowledge and leave beauty! Hello everyone, this is Xiaobian. Welcome to continue to visit the content of the academy, and we will bring you more and better content sharing wholeheartedly.

【思维导图】

【基本概念】

同一个测验项目，对同一组人员进行前后两次测试，两次测试所得分数的相关系数即再测信度，它反映两次测验结果有无变动，也就是测验分数的稳定程度，故又称为稳定性系数。

For the same test item, the same group of people are tested twice before and after, and the correlation coefficient of the scores obtained from the two tests is the retest reliability, which reflects whether the results of the two tests change or not, that is, the stability degree of the test scores, so it is also called the stability coefficient.

稳定性系数主要分为两种，Pearson相关系数和Cohen Kappa指数。

There are two kinds of stability coefficients, Pearson correlation coefficient and Cohen Kappa index.

再测信度实质是求同一量表在两次测试中的相关系数。通常是求Pearson相关系数。

The essence of retest reliability is to find the correlation coefficient of the same scale in two tests. Pearson correlation coefficient is usually found.

再测信度必须满足以下几个假设：

The retest reliability must meet the following assumptions:

1.所测量的特质必须是稳定的。

2.遗忘和练习的效果相同。

3.两次测试期间，被试者对问题的熟悉情况没有差别。

1. The measured traits must be stable.

2. Forgetting has the same effect as practicing.

3. During the two tests, there was no difference in the subjects' familiarity with the problem.

若以上三条不易做到，则不宜采用再测信度。在获得再测信度时，两次测量的时间间隔要适当。时间太短会增加稳定性，时间太长容易受到其他因素变化的影响。同时第二次测量时需要调动被试者的积极性，避免被试者对于重复测量引起不必要的情绪反应，影响第二次测试的效果。

If the above three items are not easy to achieve, retest reliability should not be adopted. When retest reliability is obtained, the time interval between two measurements should be appropriate. Too short time will increase stability, and too long time will be easily affected by changes in other factors. At the same time, it is necessary to arouse the enthusiasm of the subjects during the second measurement, so as to avoid unnecessary emotional reactions caused by repeated measurements and affect the effect of the second test.

Cohen Kappa指数，Kappa指数是用来描述两种测量手段的一致性，如果其中一种手段为标准测量手段，那么它就是标准效度。

Cohen Kappa Index and Kappa Index are used to describe the consistency of two measurement methods. If one of them is a standard measurement method, it is standard validity.

当卡方检验认为两种测量结果具有一致性后，可以进一步计算反映一致性的指标Kappa指数。我们使用Kappa指数来描述两种测量手段的一致性。根据经验，Kappa大于0.75，可认为一致性较好；Kappa大于等于0.4，小于等于0.75，说明一致性中等；Kappa小于0.4，则表明一致性较差。

When Chi-square test shows that the two measurement results are consistent, the Kappa index reflecting the consistency can be further calculated. We use Kappa index to describe the consistency of the two measurement methods. According to experience, Kappa is greater than 0.75, which can be considered as good consistency; Kappa is greater than or equal to 0.4 and less than or equal to 0.75, indicating moderate consistency; If Kappa is less than 0.4, it indicates poor consistency.

【实例分析】

接下来我们来看一个再测信度的实例分析：

Next, let's look at an example analysis of retest reliability:

心理调查第一次调查的数据如下表所示，第二次调查的数据如第二张表所示，试对该量表进行再测信度分析。

The data of the first survey of psychological survey are shown in the following table, and the data of the second survey are shown in the second table. Try to analyze the retest reliability of this scale.

第1步，分析，进行再测信度分析。

The first step is to analyze the retest reliability.

第2步，数据组织，建立“内向性”~“机动性”8个变量及这8个变量的总分“total”变量和“内向性1”~“机动性1”及这8个变量的部分“total1”外加一个序号变量，一共19个变量。

Step 2, data organization, establish 8 variables of "introversion" ~ "mobility" and the total score "total" variable of these 8 variables, and "introversion 1" ~ "mobility 1" and part of these 8 variables "total 1" plus a serial number variable, a total of 19 variables.

第3步，按下图所示，进行再测信度分析设置。

Step 3, as shown in the following figure, set up the retest reliability analysis.

第4步主要结果及分析。

Step 4: Main results and analysis.

然后是关于卡帕系数的实例分析：

Then there is an example analysis of Kappa coefficient:

两名放射科医师对200名肺病可疑患者的X光片进行图片的诊断结果如下表所示，请计算Kappa指数。

Two radiologists examined the X-rays of 200 patients with suspected lung disease. The diagnosis results are shown in the following table. Please calculate Kappa Index.

第1步，分析，由于考察的是两个医师读X光片的一致性，故可以用Kappa指数度量。

The first step, analysis, can be measured by Kappa index because it examines the consistency of X-ray films read by two doctors.

第2步，数据组织，建立“第一次检查”，“第二次检查”和“例数”三个变量，输入数据并保存。

Step 2, data organization, establish "first inspection", "second inspection" and "number of cases" three variables, input data and save.

第3步，数据加权，按“数据->个案加权”打开加权设置对话框，并将例数变量作为加权变量。

Step 3, data weighting, press "data-> case weighting" to open the weighting setting dialog box, and take the case number variable as the weighting variable.