English subtitles

← R Output - Intro to Inferential Statistics

Get Embed Code
1 Language

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

  1. In fact, the t-statistic I gave you is the one that we get when we do a
  2. hypothesis test for the plane flight distance versus cost that we talked about
  3. earlier. When we do linear regression in R, the statistical program that I've
  4. briefly mentioned before, this is the output. Usually, you'll do software to do
  5. any type of hypothesis test. Whether it's on a sample mean or two sample means
  6. or the correlation coefficient or the slope, or whatever. The important thing
  7. is that you understand the meaning of the software output. Here you see the t
  8. value for this slope is pretty big, 5.77, the same value I gave you earlier,
  9. and the p-value is pretty small. There are three stars by it, indicating that
  10. the p-value is small enough to reject the null. Here are the significance
  11. codes, which list the alpha levels. We see that three stars means we can reject
  12. the null at the significance level of point 0.001. Since this slop is
  13. significant, we know there's a true relationship between the distance traveled
  14. and the cost of the ticket. Note that we can also conduct a hypothesis test on
  15. the Y intercept, and in this case we see that's its not significant. The p
  16. value is pretty high at 0.2 and there are no starts by it. This means that
  17. we're not totally sure that the true intercept is at 161.38775. However, in
  18. many cases, the estimate for the intercept is not of interest to researchers.
  19. One reason is that this value could have no meaning in real life. Because x
  20. equals 0 may not be realistic. For example, the price of iPods at the year zero
  21. AD. iPods weren't even invented back then.