Question

2. (2   True-False. Just say whether each statement is True or False – no need to justify your answer.

1. If the number of trials in the binomial distribution increases by 1 (and P equals .50), the probability of getting either of the most extreme possible outcomes (that is, 0 or N) is cut in half.

2. If the number of trials in the binomial distribution increases by 1 (and P does not equal .50), the probability of getting either of the most extreme possible outcomes (that is, 0 or N) is cut in half.   

3. In the binomial distribution, as the probability of a “+” or “-” outcome differs more and more from .5, the shape of the distribution of probabilities for all possible outcomes always becomes more and more symmetrical.

4. In the sign test (where P = .50), the total p value associated with a given number of “+” outcomes will be the same as the p value associated with the same number of “-” outcomes, when the test is two-tailed.

5. In the sign test, when the obtained result is exactly what would be expected by chance (like 9 successes out of 18 trials when P = .50), the p value (assuming you’re doing a two-tailed test) can sometimes be greater than 1.00.

6. In the sign test, if N increases from 15 to 20 and alpha is made less stringent (like from .01 to .05), the number of distinct possible outcomes (e.g., number of heads out of N) that allow rejection of H₀ must increase.

7. In the sign test, if N decreases and the size of the effect of the independent variable decreases in strength, the probability of a Type II error decreases.

8. In the sign test, as the numerical value of P_real decreases, the power of an experiment must always decrease.

9. As power decreases, the probability of correctly rejecting the null hypothesis increases.

10. As power increases, the probability of making a Type II decreases.

11. If a researcher fails to reject the null hypothesis, then she must “accept” the null hypothesis.

12. A researcher will always know for sure when she has made a Type I error.

13. If the obtained p value is less than the alpha level, the null hypothesis should be rejected.

14. If the obtained p value is equal to the alpha level, the null hypothesis should not be rejected.

15. If the obtained p value is equal to the alpha level, the null hypothesis should be rejected.

16. It is impossible to make a Type I error when you reject the null hypothesis.

17. It is impossible to make a Type II error when you reject the null hypothesis.

18. If a researcher uses a one-tailed test, it will be easier for her to reject the null hypothesis than if she uses a two-tailed test, assuming the effect is in the predicted direction.

19. If a researcher uses a one-tailed test, it will be easier for her to reject the null hypothesis than if she uses a two-tailed test, even if the effect is not in the predicted direction.

20. If a researcher uses a one-tailed test, and the observed effect is very strong but not in the predicted direction, she must fail to reject the null hypothesis.

21. All else being equal, if the N in the sign test decreases, it becomes easier to reject the null hypothesis.

22. In a binomial distribution with 15 trials, increasing the alpha level from .01 to .05 means that fewer of the possible particular outcomes (like number of trials correct) will allow rejection of the null hypothesis.

23. All else being equal, if a researcher increases the likelihood of making a Type I error by increasing the value of alpha, then she is also more likely to make a Type II error.

24. When you have a known population and want to derive the sampling distribution of the mean from that population (like in the z test), if you vary the sample size (N), the resulting sampling distributions will have different means.

25. When you have a known population and want to derive the sampling distribution of the mean from that population (like in the z test), if you vary the sample size (N), the resulting sampling distributions will have different standard errors.

26. In the z test, as the sample size N increases, the resulting sampling distribution of the mean will become more normal in shape, regardless of the shape of the original population distribution.

27. In the z test, if you change the sample size, this will change the value of z_crit.

28. In the z test, |z_crit| at an alpha level of .05_2-tailed will always be 1.96.

29. In the z test, |z_crit| at an alpha level of .05_2-tailed will always be higher than |z_crit| at an alpha level of .01_2-tailed.

30. The computation of the effect size Cohen’s d is affected by changes the sample size in the same way that the standard error is.

Answer 1

Answerd : - True 14 - False False False True False False True False - False - False - False 20 - False False True False False - - - True 23 - False - True False True False False - False 13 - True – THE END - Scanned with CamScanner