标签： testing hypothesis-test uncertainty
如果我进行了测试，然后重复测试n=2000，则在确定名义水平alpha =0.05时，我将评估类型I错误。
n=2000
alpha =0.05
我发现我的类型错误在某些情况下有所不同，并非完全为0.05。即0.035,0.04,0.045,0.055, or 0.06？ 我们如何确定观察到的I型错误何时足够接近0.05？
0.035,0.04,0.045,0.055, or 0.06