To provide an overview
of the content areas to be covered on the final
exam:
Descriptive
statistics for categorical data (Module 5):
 Know how to compute
probabilities from a crossclassification table,
including unconditional (marginal and intersection)
and conditional (row and column probabilities).
 Inferences applied to
single proportions
 For an hypothesis
test, be able to set up the null and alternative
hypotheses, select an appropriate pvalue from
computer output, and make the three part test
conclusion
 Be able to
interpret a confidence interval from computer
output; know how to narrow or widen the margin of
error, and know how to compute sample size.
 Inferences applied
to difference between two proportions
 For an hypothesis
test, be able to set up the null and alternative
hypotheses, select an appropriate pvalue from
computer output, and make the three part test
conclusion
 Inferences applied
to difference between c proportions using
crosssclassification tables and ChiSquare
Statistics
 From computer
output, know how to interpret the
crossclassification table, test the hypothesis
that the proportions are not equal, test the
hypothesis that two variables are related, and
interpret Cramer's Phi Coefficient.
Model Building with
Regression Analysis (Modules 2 and 3)
 On the front end:
identify experimental unit (object, person, or thing
upon which measurements are taken); dependent
variable, independent variable; and population,
sample, representative sample (if sample was randomly
selected or not)
 Model identification:
from graph or hypothesized regression equation,
identify curvature model with and without interaction,
linear model with and without interaction, and model
with just categorical dummy variable
 Output
interpretation: Interpret regression coefficients (the
B's) according to model selected; determine practical
utility (standard error and R Square); determine
statistical utility with appropriate hypothesis test;
interpret presence/absence of curvature, interaction,
categorical dummy variable, and/or quantitative
variable.
 Construct and
interpret confidence interval for predicting y.
 Determine if model
assumptions (error is normal; error has constant
variable) are met
Descriptive and inferential statistics
on quantitative data (Module 1) and Comparison of Means
(Module 4)
 From computer output, interpret
descriptive statistics (mean, standard deviation,
median, interquartile range, left/right skew, normal
or mound shaped distribution, outliers, and voice of
process (mean plus/minus three standard
deviations)
 Be able to state and interpret
null and alternative hypotheses, and using computer
output, make three part test conclusion for testing if
one mean is equal, less than or greater than some
number; testing if two means are equal or one is less
than/greater than another mean; testing if multiple
means are equal or not using ANOVA and if they are not
equal, determining which pairs are different.
