What is the purpose of inferential statistics? How is it different than descriptive statistics? Are they complementary or competitive?
Why do we care about probability at all when we are talking about inferential statistics?
What are the four steps involved in hypothesis testing?
What is meant by specifying the null hypothesis? What about the alternative hypothesis? What is the benefit of being so specific about the hypotheses?
As discussed in class, what are the differences between the "classical" and "contemporary" approaches to hypothesis testing?
With inferential statistics, the goal is to reject the null hypothesis. What does this mean? Do we conclude that the alternative hypothesis is correct? Why or why not?
Why is the standard error of the mean (based on many samples) going to be smaller than the standard deviation of a single sample? In explaining your answer, be sure to describe the interpretation of a standard error of the mean.
What types of error can occur when making decisions based on inference? Be specific.
Why are points beyond + or - 2 standard deviations often considered outliers by many researchers?
What does it mean if a researcher sets his/her alpha at .01, and rejects the null hypothesis? How does this differ from setting the alpha at .05 and rejecting the null? In which case is the researcher going to be most likely to reject the null hypothesis?
What are degrees of freedom, and when do you use them?
What is the difference between a one-tailed and a two-tailed test? When would you use each (if at all)?
What is meant (and not meant) when a researcer says a finding is "statistically significant"?
What is the generic formula for the T-Test? Explain it conceptually (not mathematically).
Why would we consider using the Mann-Whitney or Wilcoxon tests, rather than the more traditional T-Tests?
What is the relationship between T-Tests and F-Tests?
What is the generic formula for the F-Test? Explain it conceptually (not mathematically). How does this relate to the generic formula for the T-Test?