Brannick, Spring 2002
Exam 1
- What
happens to the mean, median, and mode when the distribution becomes
skewed? Draw a picture to
illustrate your answer.
- What
is a sampling distribution?
- Describe
the standard error of the mean,
, when samples are drawn from a normal
distribution. What determines its
size?
- Why is
the number 1.96 such a big deal for the normal distribution?
- How
does the null hypothesis help specify the rejection region?
- What
is statistical power? Why is it
important?
- How
are the distributions of z and t related?
- What
are the mean and variance of the chi-square distribution?
- Describe
the F distribution. What is
it made of? What affects its shape
(what are the parameter(s) of the F distribution)?
- Give a
concrete example of the use of the dependent samples t-test. State why the particular test is the
right one to choose, that is, what it is about the design that makes the
dependent samples t appropriate.
Exam 2
- What
is the difference between fixed and random factors in ANOVA? Give an example of each.
- Why is
the sum of squares for error important in ANOVA significance tests?
- Describe
(make up) a concrete example of a one-way ANOVA where it makes sense to
use an overall F test. Explain why
ANOVA (not t, chi-square or something else) is the best method for the
analysis.
- Describe
(make up) a concrete example where you would use a post hoc test. Explain why the post hoc test is needed
(not the specific choice of post hoc test, but rather why post hoc test at
all).
- Define
(label) each term in this linear model:
-
Describe a concrete example of a randomized block design. You should have 1 factor as the
blocking factor and one other factor as the factor of main interest.
- What
is the purpose of the Fisher r to z transformation?
- What
is the relation between r in correlation and R-square in regression with 1
independent variable?
- What
is the problem in regression/ANOVA if the homogeneity of variance /
homoscedasticity assumption is violated?
What will happen in practice?
10. What is the difference between a per comparison and a familywise
error rate?
Exam 3
- What
is the difference between a standardized and unstandardized regression (b)
weight? Why (under what
circumstances) might you prefer one to the other?
- Why is
regression more closely related to the semipartial correlation than to the
partial correlation?
- Draw
an illustration of the confidence interval for the regression line. Why does it have the shape that it
does?
- What
is cross validation? How does it
relate to the ‘adjusted R-square’ contained in the SAS regression output?
- Why
does the order of entry into a hierarchical regression change the
increment in variance due to a particular variable?
- What
is the variance inflation factor (VIF)?
What is its major use in regression analysis?
- Suppose
you were predicting employee salary based on the financial responsibility
of the job (that is, the yearly salary for a job is predicted by the size
of the budget controlled by the job).
Upon looking at the residuals of the linear regression, you notice
that salary for jobs at the high budget end are under-predicted. How can you solve this problem? Describe how you would modify the regression
equation.
- Why
should you avoid dichotomizing continuous variables for analysis (describe
2 reasons)?
- What
are least squares means? Describe
a concrete situation in which you might want to use them.
- What
is the connection between homogeneity of variance in ANOVA and
homoscedasticity in regression?
- Suppose
you have two unbiased statistics, A and B. A has a standard error that is twice that of B. Which statistic are you likely to
prefer and why?
- What
is the difference between the t distribution and the unit normal (z)? When do we prefer to use t to z?
- Why do
we examine studentized residuals in regression? What do they help us to do?
- We
could use the t-test for comparing single means or ANOVA to test groups of
means in post hoc test. Why don’t
we?
15. Describe the chi-square distribution. What is it made of?
Describe its shape. What is its
expected mean value?