Basic Applied Statistics 200
Extra Credit Assignments
Maximum of 5 points each, total maximum of 50 points; hand any of these in to
me [Nancy Pfenning] by April 16. I strongly urge you to start early and hand
in just one or two problems the first time, in case you are taking the wrong
approach and need some guidance.
Use any of the data provided in the compiled responses to our class
survey. Be sure to begin each problem with an opening statement, telling what
you intend to do. Tell what you anticipate the outcome of your analysis will
be. (If you really can't make a guess, just say so.) Finish with a
concluding statement which summarizes your findings.
Remember, I'M giving YOU the assignment, not the other way around! If you
use MINITAB, all
output must be explained by YOU in order to receive credit.
You may work together if you like, but (a) make a note of this to me; and
(b) all explanations must be in your OWN words.
- Thoroughly describe and display a quantitative data set. Be
sure to mention center, spread, shape, and possible outliers;
include stemplot, boxplot, and histogram.
- Compare values of a quantitative variable for 2 or more groups. Be
sure to compare centers, spreads, and shapes; include back-to-back
stemplots and/or side-by-side boxplots.
- Examine the relationship between 2 quantitative variables. Include
a scatterplot, mention of direction, form, and strength; correlation and
the regression line equation if the relationship appears linear; mention
of outliers or influential observations if present.
- Examine the relationship between 2 categorical variables. Display
the data with a two-way table; show marginal and conditional
distributions, and display the data with bar graphs. Summarize the
relationship by comparing the most relevant percentages.
Note: for the statistical inference problems, we will assume our class
to be a random sample of Pitt students. This may or may not be a
reasonable assumption, depending on the variables studied. Unless you
happen to know population standard deviation (eg. for heights or
weights), you will need to work with sample standard deviation. This
will affect your choice of a z or t procedure. Examine
plots of the data to decide if the Central Limit Theorem applies for the
given sample size.
- Set up a confidence interval for the mean of a quantitative variable.
- Test a hypothesis about the mean of a quantitative variable.
- Set up a confidence interval for the difference between means for 2
groups. (Look at sample standard deviations to decide whether or not to
use a pooled procedure.)
- Test if the difference between means for 2 groups is zero. (Decide
whether or not to use a pooled procedure, and make sure you formulate a
reasonable alternative hypothesis.)
- Set up a confidence interval for a proportion based on categorical data.
- Test a hypothesis about a proportion.
- Set up a confidence interval for the difference between proportions
for 2 groups.
- Test if the difference between proportions for 2 groups is zero.
- Use chi square to test for a relationship between 2 categorical variables.
- Use ANOVA to compare means for more than 2 groups.
- Examine a relationship between 2 quantitative variables as in #3.
If it appears linear, estimate the regression model parameters and tell
what they mean;
test the null hypothesis that the slope of the regression line is zero.
[ Home
| Calendar
| Assignments
| Handouts
]
|