知识卡片
子样本测试检查稳定性
随机减少部分样本、重复估计并比较分布,可以检查结果对样本组成与规模是否敏感。稳定支持这类扰动下的稳健性,不证明剩余数据已足够,也不排除所有子样本都共享同一种因果偏差。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
原书出处 · 1 处
- · 查看原书依据
11.5.1 Data size reduction One way to test the robustness of the analysis is to reduce the size of the data and see if we obtain similar results. We are assuming our analysis has more than enough data to achieve a stable estimation. We can refute this assumption by slightly reducing the size of the data and testing whether we get a similar estimate. Let’s
参考来源
- 位置:《Causal AI》第11章《构建因果推断工作流》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-11.xhtml)
- 结论依据:摘录中关于子样本测试检查稳定性的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:11.5.1 Data size reduction One way to test the robustness of the analysis is to reduce the size of the data and see if we obtain similar results. We are assuming our analysis has more than enough data to achieve a stable estimation. We can refute this assumption by slightly reducing the size of the data and testing whether we get a similar estimate. Let’s