进阶专项：研究设计与证据强度

唯一新困难：证据强弱

不同研究设计支持的结论强度不同。个案说明“可能存在”，相关观察支持“同时变化”，对照试验才更接近因果。阅读不要求统计学细节，但必须让选项语气匹配证据等级。

自拟短文

A survey of 600 commuters found that people who read on the train reported lower stress than those who checked work messages. Because participants chose their own activity, the survey cannot show that reading itself caused the difference. It does, however, identify a pattern worth testing experimentally.

题目：What conclusion is justified?

A. Reading on trains definitely reduces stress.
B. Work messages are the sole cause of commuter stress.
C. Reading and lower reported stress were associated in the survey.
D. The experiment proved the pattern false.

证据等级

这是自选活动的调查，能报告关联，不能排除本来压力较低的人更爱阅读等替代解释。因此 C。A 把关联加强为确定因果；B 添加唯一原因；D 把“值得实验检验”偷换成实验已完成且否定。

措辞匹配

观察证据适合 was associated with / was linked to / coincided with；因果措辞 caused / led to / resulted in 需要更强设计。prove 通常比 suggest 要求更高。

独立练习

哪种后续研究更能检验因果？

**答案：**将可比参与者随机分配为阅读与查看工作消息，并用统一方式测量压力，同时控制乘车时长等条件。

独立动作验收

写出本研究设计、可支持结论、不可支持结论和一个替代解释。若仍使用 definitely 或 sole，视为证据强度失配。