Hi there!
I am currently conducting a hypothesis test and I am unsure which tailed test would help support my claim. I am testing median income and poverty rates. I split the group into above and below the median income, giving me a rich and poor group. The null hypothesis is mean of rich = mean of poor.
I am trying to support the claim that as your median income increases, poverty rates should fall. I am confused as they are inversely related and unsure as to what the alternative hypothesis should be. And whether I should conduct a left or right or both tailed test. Any help would be appreciated!
Thanks!!
The tailed (t?) tests are useful if you have a more specific question (e. g. ¿Does the "rich" group has a higher income than the "poor" group?). So, the left or right-tailed test depends on your question and how it is formulated, so you should be careful in the way you interpret the results.
Take a look on this link: http://www.r-tutor.com/elementary-statistics/hypothesis-testing/two-tailed-test-population-proportion
Hello Trung,
From your explanation, it sounds as if a directional hypothesis test is more appropriate for you (though the question is a bit of a tautology: If you're richer, you're less likely to be in poverty).
So, Ho is: Richer folks have equal or higher level of poverty than poorer folks.
Ha is: Richer folks have lower level of poverty than poorer folks.
Is there a reason that you couldn't just correlate the two variables, instead of dichotomizing income into richer/poorer?
Good luck with your work.
I agree with David Morse that your statement is tautological. Is the income part of the definition of poverty rate? Examine the possible conclusions and their usefulness before investing your effort in obvious results.
I join the proposal of relating the variables, drawing graphs to explore the type of relationship (linear or not) between them. You may want to use Spearman or Kendall coefficients, or choose generalized models.
- Badges
- Science topic
- Similar topics
- Political Science