**This answer sheet is for the final exam that says "Final exam Fall 2012" at the top, with no comma after "Final exam." If your exam says "Final exam, Fall 2012" with a comma go to the other answer sheet.**

- Number of years: measurement. Volume of damage, measurement. Linear regression/correlation.
- Marsh, nominal. Before or after restoration, nominal. Number of ducks, measurement. Paired t-test (=two-way anova, no replication).
- Steroid, placebo or no treatment, nominal, meaningful. Arm strength, measurement. One-way anova, model I. (Note that the question does not say that you measured arm strength before the treatment).
- Person's house, nominal. Copper wire or not, nominal. Number of mosquito larvae, measurement. This would normally be analyzed with a paired t-test (=two-way anova with no replication), but the histogram of differences between copper and no copper is not normally distributed, so Wilcoxon signed-rank test.
- Age of bird, nominal. Identity of each bird, nominal (because you have multiple measurements per bird). Number of calls, measurement. Nested anova.
- Amount of light, measurement. Cell length, measurement. Linear regression/correlation.
- The first step in a meta-analysis is to set the criteria for which papers you will include in your meta-analysis. You should do this before you start collecting papers, so you're not tempted to set your criteria to exclude papers whose results you don't like.
- Carbon dioxide concentration, 390 or 600 parts per million, nominal, interesting. Size, measurement. One-way anova, model I (=Student's t-test).
- The cancer cells could have a smaller sample size or a larger standard deviation.
- Type of exercise, nominal, interesting. Amount of protein, measurement. One-way anova, model I.
- Right knee or left knee, nominal. Right or left-handed, nominal. Small sample size. Fisher's exact test.
- Mutant type, nominal, interesting. Pupation height, measurement. One-way anova, model I.
- Warmup time, measurement, independent. Water amount, measurement, independent. Stretching time, measurement, independent. Time in 100 meters, measurement, dependent. Multiple regression.
- Diet, nominal, interesting. Appendix size, measurement. One-way anova, model I.
- Light vs. dark, nominal. Rat name, nominal (you have five measurements per rat). Clarity, measurement. Nested anova.
- Size of cell, measurement. Number of chloroplasts, measurement. Linear regression/correlation.
- Distance from end of tank, measurement. Length of fish, measurement. Linear regression/correlation.
- The dollar sign tells SAS that "gene" is a nominal variable. (Just saying "It indicates the nominal variable" wasn't enough, since that doesn't show that you know whether "gene" or "fst" is the nominal variable).
- Type of bone cell, nominal. Intensity of green, rank. Kruskal-Wallis test.
- Age, height, weight, stride length, all measurement, all independent. Joint space, measurement, dependent. Multiple regression.
- Volunteer name, nominal, random and uninteresting. Blood pressure, measurement. One-way anova, model II.
- Point being stared at, nominal. Name of person, nominal (because you have multiple measurements per person). Standing time, measurement. Two-way anova, no replication.
- Cell growth is signficantly different among the three different brands of culture medium, and cell growth is signficantly different among the four cell lines. Because the interaction term is significant, the effect of the brand of culture medium depends on which cell line you're looking at.
- Amount of moisture, measurement. Egg turns into live fly or not, nominal. Logistic regression.
- Identity of tree, nominal, random and uninteresting. Size of leaf, measurement. One-way anova, model II.
- Type of singer, nominal. Size of uvula, measurement, not normally distributed. Kruskal-Wallis test.
- Quadrant each amphipod is stuck in, nominal. Theoretical expectation (equal numbers in each quadrant). Large sample size (1500 total). Chi square or G-test of goodness-of-fit.
- Baby or not, nominal, interesting. Identity of mother, nominal (because you have multiple measurements per mother). Length of jump, measurement. Nested anova.
- Annual income, measurement. Dental health, measurement (they're not ranked in order, dental health is measured using a subjective 100-point scale). Linear regression/correlation.
- Type of soil, nominal, interesting. Number of apples per tree, measurement. One-way anova, model I.
- Beta (or power, which is 1-beta) and standard deviation of polar bear weight.
- Arsenic concentration, measurement. Number of Halomonadaceae, measurement. Linear regression/correlation.
- Type of cover crop, nominal, interesting. Number of corn borers per 10 m
^{2}section, measurement. One-way anova, model I. - Hat or no hat, nominal. Baldness area, measurement. Age, measurement. Ancova.
- French fry or minnow, nominal. Ring-billed or herring gull, nominal. Small sample size. Fisher's exact test.
- Cell line, nominal. Dish, nominal (because you have multiple measurements per dish). Sodium level per cell, measurement. Nested anova.

Return to the Biological Data Analysis syllabus

Return to John McDonald's home page

This page was last revised December 13, 2012. Its URL is http://udel.edu/~mcdonald/statexam3comma.html