Friday, February 5, 2010

The probability of rejecting a true null hypothesis increases as the sample size increases?

Are you asking if this is true? It is. A null hypothesis is one that proves there is no relationship or a relationship between two things is not true. You can never really prove a hypothesis to be true; you can only reject the null hypothesis, that is, reject the fact that is not true. An example: say you are studying the size of amazonian bees in the rainforest and want to know if the size of the bees is different on the west side vs the east side (a hypothetical exp of course). Null hypothesis would be that there is no difference. A different hypothesis would be that either west are larger or east are larger.





The proability of rejecting a true null hypothesis increases with sample size because you have a much better representation of your samples, or a better representation of the whole forest. If you sampled only 5 bees,say, and 3 were large and 2 were small on teh west side, you would say that you cannot reject the null hypothesis b/c there actually is a difference, there are more larger ones.





However, 5 bees is a small sample size and not indicative of the true population of bees. A larger sample makes it easier to show teh diversity, true population, and thus it is easier to reject a real null hypothesis. Hope this helps!The probability of rejecting a true null hypothesis increases as the sample size increases?
I'm going to back away from this one. I started to answer thinking I knew the answer, but now I am not sure. Sorry.

No comments:

Post a Comment