SPSS

Full course: https://www.youtube.com/watch?v=ZpwZS3XnEZA#t=8m41s 

samples: https://www.youtube.com/watch?v=ZpwZS3XnEZA#t=16m22s c:program filesibmspsssp26samplesdemo.sav 18m50s:Descriptive Stat|Explore 19m36s:basic graph, 20m48s(https://drive.google.com/drive/folders/1LCtTh6Gt2PqFSLlUt_z2Mhu3Ha-KFujT)

0. Original papers/Installation

网上找老友的spss 26.zip，与Python 3.4一起装在W11下，sample下例子还没学。复原了这两篇文章的结果：线性混合效应模型入门之一（linear mixed effects model）线性混合效应模型入门之二 原始数据：https://pan.baidu.com/s/1PYcMxDYgjjSvv9seUXmIDg?pwd=ngkw

SPSS 26 documents：https://www.ibm.com/docs/en/spss-statistics/26.0.0

本案例数据来源于一个肾脏病的研究。研究对200个肾病患者进行随访，每年化验一次肾小球滤过率（GFR，评价肾脏功能的指标，会逐年下降）。主要分析目的是探索基线的尿蛋白定量对GFR年下降率（斜率）的影响（尿蛋白量越大，对肾功能危害越大），混杂因素包括基线年龄和性别。

The data in this case come from a study on kidney disease. The study followed 200 patients with kidney disease and tested the glomerular filtration rate (GFR, an indicator of kidney function, which decreases year by year) once a year. The main purpose of the analysis was to explore the impact of baseline urinary protein quantification on the annual decline rate (slope) of GFR (the greater the amount of urinary protein, the greater the harm to renal function). Confounding factors include baseline age and gender.

字段说明：
（1）patient: 患者ID编号；
（2）visit：化验次序编号；
（3）time：化验时间（单位年），第一次化验定为0，后面依次推延；
（4）GFR：肾小球滤过率，单位是ml/min/1.73^2，作为响应变量；
（5）age：基线年龄，单位岁；
（6）gender：性别，0=男，1=女；
（7）micro：基线是否有微量蛋白尿，0=无，1=有；
（8）macro：基线是否有大量蛋白尿，0=无，1=有；
补充说明：
（1）蛋白尿这里用了哑变量编码，macro=0且micro=0表示没有蛋白尿；
（2）数据中GFR化验数据有缺失，线性混合效应模型对缺失数据有良好的处理能力。

1. SPSS start note

1. Rub SPSS, close welcome window, operate on main GUI.
2. File|Import Data|CSV Data, Open .csv file, then click OK

3. Analyze|Mixed Models|Linear,

4. Set model, start by choose patient from left, use arraw to add to right pane, Continue (below left)

5. Add factors, as shown at above right, click Fixed.
6. Add fixed effect as shown above. Select multiple params for interaction *. Click Continue.

7. Set Covariance Type and other factors as shown below left. Click Continue.

 

8. Click Estimation, change method as shown above right, then click Continue.
9. Click Statistics, then Save, both with Continue after settings, shown below, then OK!

 

10. Outputs items shown at the light and details at the right, for example:

11. Click Disk symbol to save it: