Talk:Statistical hypothesis test
Hello. I've reworked the opening to state a definition for "statistical hypothesis test". Hopefully it's an improvement over the previous, which stated only:
- Many researchers wish to test a statistical hypothesis with their data.
Feel free to further improve it. -- BTW I see the article has a strong Karl "induction is impossible" Popper bias; I don't think that's necessary, even within the realm of frequentist probability. It would be interesting to trace the history of hypothesis testing as an implementation of scientific method; I don't know what Fisher, Neyman, & Pearson, Wald, etc., said about that. -- I'm aware that a distinction is made between "significance tests" and "hypothesis tests"; I guess that distinction should be clarified in this article. I don't know if separate articles are needed. Happy edits! Wile E. Heresiarch 14:51, 17 Feb 2004 (UTC)
a tyre manufacturing company`s sales manager claims that all the tyres produced by a company have a tread life of at least 5000 kilometres,64 tyres are sampledfrom a batch of the tyres and the tread life mean of the sample is found to be 8000kilometres.The standard deviation of the production of the tyres is 4000 kilometres.Can you call the company`s sales manager an impostor based on the sample?Assume 5%level of significance and the normal distribution of the tread life of the tyres in a two-tail test?
- and do you mean standard deviation of the sample? or of the entire production?
- in all practicality, perhaps not an imposter, but the sales manager should certainly warranty replacement of any tyres that do not meet the claimed performance;
- In other words, we're not going to do your homework for you. :-)
The following appeared on the page but looks like a comment from User:Ted Dunning
- Note: Statistics cannot "find the truth", but it can approximate it. The argument for the maximum likelihood principle illustrates this -- TedDunning
Should this article perhaps be filed under Statistical test or Statistical hypothesis test with Hypothesis testing, Testing statistical hypotheses and Statistical hypothesis testing all #REDIRECTing to it? Oh, and is the "to-do list" box at the top of this page really necessary? If so, maybe it should be explained a little more. It took me a while to understand its significance (no pun intended). In particular, the phrase "Here is" is ambiguous, and even misleading, since it's often used to refer to the links that follow. - dcljr 07:54, 6 Aug 2004 (UTC)