如何让SPSS中的DO REPEAT语法正常运行？

核心错误：DO REPEAT变量组数量不匹配 #

DO REPEAT要求每组循环变量的数量完全一致，但你的语法中变量组数量差异明显：

• AE_trial_numberedN1到AE_trial_numberedN134：134个变量
• Start_med1到Start_med39：39个变量
• i_aestdt1到i_aestdt134：134个变量
• i_aeendt1到i_aeendt134：134个变量
• ABused1到ABused39：39个变量
• AB_AE_reln1到AB_AE_reln134：134个变量

SPSS会按位置一一配对循环变量，数量不匹配会直接导致执行失败。你需要确认变量的对应逻辑：是AE和用药记录的多对多关系？还是调整变量组的范围，让每组数量一致。

次要语法错误 #

1. 逻辑条件优先级错误
   最后两个else if中，OR连接的AE_trial_numberedN多值判断未加括号，由于&优先级高于|，会导致逻辑解析错误。正确写法应把多值判断用括号包裹：

else if ((AE_trial_numberedN=7 | AE_trial_numberedN=8 | ... | AE_trial_numberedN=22) & start_med > i_aestdt & start_med < i_aeendt and ABused > 0).

2. 语句结尾缺少句号
   END REPEAT语句末尾必须加句号，SPSS语法要求每个完整语句以句号结尾：

end repeat.

3. 变量名大小写不一致
   代码中同时出现Start_med和start_med，虽然SPSS默认不区分大小写，但统一变量名可避免潜在混淆。

修正后的示例语法（需根据实际数据调整变量范围） #

* 假设调整为每组变量数量一致，示例逻辑修正，需匹配你的数据结构.
do repeat AE_trial_numberedN = AE_trial_numberedN1 to AE_trial_numberedN39 
/ Start_med = Start_med1 to Start_med39 
/ i_aestdt = i_aestdt1 to i_aestdt39 
/ i_aeendt = i_aeendt1 to i_aeendt39 
/ ABused= ABused1 to ABused39 
/ AB_AE_reln = AB_AE_reln1 to AB_AE_reln39.

    do if (AE_trial_numberedN=1 & Start_med > i_aestdt & Start_med < i_aeendt and ABused > 0).  
    compute AB_AE_reln=1.  
    
    else if (AE_trial_numberedN=1 & Start_med > i_aestdt & Start_med < i_aeendt and ABused = 0).
    compute AB_AE_reln=11.  
    
    else if (AE_trial_numberedN=2 & Start_med > i_aestdt & Start_med < i_aeendt and ABused > 0).
    compute AB_AE_reln=2.

    else if (AE_trial_numberedN=2 & Start_med > i_aestdt & Start_med < i_aeendt and ABused = 0).
    compute AB_AE_reln=12.

    else if (AE_trial_numberedN=3 & Start_med > i_aestdt & Start_med < i_aeendt and ABused > 0).
    compute AB_AE_reln=3.

    else if (AE_trial_numberedN=3 & Start_med > i_aestdt & Start_med < i_aeendt and ABused = 0).
    compute AB_AE_reln=13.

    else if (AE_trial_numberedN=4 & Start_med > i_aestdt & Start_med < i_aeendt and ABused > 0).
    compute AB_AE_reln=4.

    else if (AE_trial_numberedN=4 & Start_med > i_aestdt & Start_med < i_aeendt and ABused = 0).
    compute AB_AE_reln=14.

    else if (AE_trial_numberedN=5 & Start_med > i_aestdt & Start_med < i_aeendt and ABused > 0).
    compute AB_AE_reln=5.

    else if (AE_trial_numberedN=5 & Start_med > i_aestdt & Start_med < i_aeendt and ABused = 0).
    compute AB_AE_reln=15.

    else if (AE_trial_numberedN=6 & Start_med > i_aestdt & Start_med < i_aeendt and ABused > 0).
    compute AB_AE_reln=6.

    else if (AE_trial_numberedN=6 & Start_med > i_aestdt & Start_med < i_aeendt and ABused = 0).
    compute AB_AE_reln=16.

    else if ((AE_trial_numberedN=7 | AE_trial_numberedN=8 | AE_trial_numberedN=9 | AE_trial_numberedN=10 | AE_trial_numberedN=11 | AE_trial_numberedN=12 | AE_trial_numberedN=13 | AE_trial_numberedN=14 | AE_trial_numberedN=15 | AE_trial_numberedN=16 | AE_trial_numberedN=17 | AE_trial_numberedN=18 | AE_trial_numberedN=19 | AE_trial_numberedN=20 | AE_trial_numberedN=21 | AE_trial_numberedN=22) & Start_med > i_aestdt & Start_med < i_aeendt and ABused > 0).
    compute AB_AE_reln=7.

    else if ((AE_trial_numberedN=7 | AE_trial_numberedN=8 | AE_trial_numberedN=9 | AE_trial_numberedN=10 | AE_trial_numberedN=11 | AE_trial_numberedN=12 | AE_trial_numberedN=13 | AE_trial_numberedN=14 | AE_trial_numberedN=15 | AE_trial_numberedN=16 | AE_trial_numberedN=17 | AE_trial_numberedN=18 | AE_trial_numberedN=19 | AE_trial_numberedN=20 | AE_trial_numberedN=21 | AE_trial_numberedN=22) & Start_med > i_aestdt & Start_med < i_aeendt and ABused = 0).
    compute AB_AE_reln=17.

    end if.
end repeat.

额外建议 #

• 若AE和用药记录是多对多关系，DO REPEAT可能不适用，建议先通过VARSTOCASES将数据转换为长格式，再进行逻辑判断，会更清晰不易出错。
• 测试语法时可先选取少量样本，逐步调试每个条件的执行结果。

内容的提问来源于stack exchange，提问作者Sven221