English 字幕

← Extended t-Test Numerator - Intro to Inferential Statistics



Showing Revision 5 created 05/25/2016 by Udacity Robot.

  1. We definitely don't want to conduct that many t-tests but we can use the same
  2. ideas underlying t-tests to compare three or more samples. Remember with
  3. t-tests, the decision of whether or not two samples are significantly different
  4. is a function of the distance that they are apart from each other. And the
  5. variability of each sample, which we call the standard error. Remember that we
  6. found this with the pooled variance. When we compare three or more samples, we
  7. can do almost the same thing. We have some kind of distance or variability
  8. between means in the numerator, and some kind of error in the denominator.
  9. Let's focus on the numerator for a sec. How can we compare three or more
  10. samples? Read each of these options carefully, and as a hint, think back to how
  11. you learned to calculate the standard deviation in lesson four. And the
  12. rationale for using the standard deviation as the measure of variability for a
  13. sample. Also, in this case, total mean is the mean of all values in all the
  14. samples.