Consider a national survey that 37% of people name soccer as their favorite sport. My personal survey says that out of 400 people, 168 say soccer is their favorite.
I'm having trouble deciding if the alternative hypothesis is:
H1 ≠ .37 or H1 > .37
Consider a national survey that 37% of people name soccer as their favorite sport. My personal survey says that out of 400 people, 168 say soccer is their favorite.
I'm having trouble deciding if the alternative hypothesis is:
H1 ≠ .37 or H1 > .37
Copyright © 2021 JogjaFile Inc.
The alternative hypothesis helps you think about what counts as extreme results under the null hypothesis.
If the national survey gave a correct result you might have expected about $148$ soccer responses to your personal survey. Would you regard $300$ as an extreme result? What about $30$?
If you care both about extremely high and extremely low, then this might suggest $H_1: \,\not = 0.37$ is the more appropriate alternative hypothesis