In: Statistics and Probability
2. Return to decision you made when answering part D of Question 1.
A. If the alpha level had been .01 (or 1%) rather than .05 (or 5%), would your decision have been different? Please explain.
B. If the alpha level had been .10 (or 10%) rather than .05 (or 5%), would your decision have been different? Please explain.
3. Return again to Question 1. Suppose that rather than setting up a one-sided (or one-tailed) alternative hypothesis, you had set up a two-sided (or two-tailed) alterative hypothesis because you were not sure if the sample results would be greater than or less than the claimed population value. Would your p-value have been different if you had done this? Please explain.
It would purely depend on the p value... If the pvalue is less than 0.05 we will reject the null hypothesis at 5% significance level. If otherwise we will not reject the null at 5% significance level
So if we use the significance level 1%.. The conclusion will be taken using 0.01ie if p value is further less than 0.01 then rejection of null would be done even at 1% level of significance.
The same can be concluded if we used 10%significance level.
If a null hypothesis is rejected at 0.01 significance level... The the same hypothesis can be rejected at higher level of significance like 5%, 10% etc
Yes p value will definitely be different. p value for two tailed test will be twice that of the one sided test. Whether the test is one tailed or two tailed, the test statistic value will remain the same... But in one tailed only one tail of the distribution of the statistic is considered... For eg if test is right tailed... Ie if the alternative is mu>mu_0.. The p value is the area towards the right side of the test statistic value in the normal curve for z test... But in the case of two tailed test both the tails will be taken in to account and area towards the left side of the - test statistic value plus area to the right side of the +test statistic value gives the pvalue... So it is double that of the one sided test...