Lecture 13. Confidence Intervals, Hypothesis Testing, and Power Calculations

preview_player
Показать описание
MIT 14.310x Data Analysis for Social Scientists, Spring 2023
Instructor: Sara Ellison

In this video, Sara Ellison discusses inference, including the topics of confidence intervals and hypothesis testing.

License: Creative Commons BY-NC-SA

Рекомендации по теме
Комментарии
Автор

10:28 this is basically the reason why one can't say that the true value is within confidence internal 95% of time, for example.

janasandeep
Автор

I think there's a typo in the slide on confidence intervals shown at time 36:39. The slide says
x_bar + phi_inverse(alpha/2) * sigma / square_root(n) < mu < x_bar - phi_inverse(alpha/2) * sigma / square_root(n)

But I think it should say
x_bar - phi_inverse(alpha/2) * sigma / square_root(n) < mu < x_bar + phi_inverse(alpha/2) * sigma / square_root(n)

That is, the SUM of x_bar and phi_inverse(alpha/2) should be on the right, and the DIFFERENCE should be on the left. I see basically the same issue in the slide at time 42:40.

Hope it helps.

NathanLGrossman
welcome to shbcf.ru