Final exam study guide: Practice exam 3 answers

Here are the answers to practice exam 1. The answer is in bold; my explanation, which you don't need to write on the exam, is in regular type.

Where two correct answers are shown for these practice questions (such as chi-squared or G-test of independence), you must only write down one on the actual exam.

  1. One measurement variable, years of football; one nominal variable, encephelopathy or not; if there's a cause-and-effect relationship, football causes encephalopathy: simple logistic regression
  2. Three nominal variables, black ducks vs. other ducks; before or after restoration; identity of marsh (marsh is a nominal variable because two sets of observations of black duck proportion per marsh): Cochran-Mantel-Haenszel test
  3. One measurement variable, arm strength; one nominal variable, steroid vs. placebo vs. nothing: one-way anova
  4. One measurement variable, number of mosquitoes; two nominal variables, copper vs. no copper, identity of yard (because two observations of mosquito number in each yard): paired t-test or two-way anova without replication
  5. One measurement variable, number of calls; two nominal variables, color of stuffed bird, identity of live bird (live bird is a nominal variable because multiple measurements per bird); each live bird found in combination with each color of stuffed bird; one measurement per combination: two-way anova without replication
  6. One measurement variable, light amount; one nominal variable, interphase or not; if there's a relationship, light amount affects probability of being in interphase: simple logistic regression
  7. Decide on the rules that will determine which studies you will include in your meta-analysis.
  8. One measurement variable, stoma size; two nominal variables, carbon dioxide concentration (nominal because just two values), individual stoma (because two measurements per stoma): paired t-test or two-way anova without replication
  9. Cancer cells could have a larger standard deviation, or you could have a smaller number of cancer cells.
  10. Check the normality and homoscedasticity of C-reactive protein level.
  11. One nominal variable, right vs. left; theoretical expectation of 1:1 ratio if null is true; sample size is less than 1000: exact test of goodness-of-fit
  12. Two nominal variables, pupa on surface or side, mutation type; total sample greater than 1000: chi-square or G-test of independence
  13. Two measurement variables, warmup distance, speed in 100 meters; relationship is curved (question says fastest speeds at intermediate warmup distances): polynomial regression
  14. One nominal variable (food type) and one measurement variable (appendix size), so the appropriate statistical test is a one-way anova; to compare one pair of categories in a one-way anova: Tukey-Kramer test
  15. One measurement variable, lens clarity; two nominal variables, identity of mouse, eyepatch vs. no eyepatch; multiple measurements per eye: two-way anova with replication
  16. One measurement variable, number of chloroplasts; one ranked variable, size of cell: Spearman rank correlation
  17. One measurement variable, length of fish; one nominal variable, each scoop of the net: one-way anova. This is an example of a "model II" or "random effects" one-way anova; you are interested in the proportion of variation among netfuls, you don't care which netful had the biggest fish.
  18. One nominal variable, type of cell; one measurement variable, fluorescence intensity: one-way anova or two-sample t-test
  19. Three measurement variables, miles run, age, and weight; one nominal variable, osteoarthritis or not: multiple logistic regression
  20. One measurement variable, blood pressure; two nominal variables, before and after Frogger, identity of person; each person in combination with before and after; multiple measurements for each combination: two-way anova with replication
  21. Two nominal variables, spot that is stared at, identity of person; each person stares at each spot; only one measurement for each combination of person with spot: two-way anova without replication
  22. Culture medium has a significant effect on cell growth, cell line does not have a significant effect, and the effect of culture medium is not different for different cell lines.
  23. Two nominal variables, moisture amount (nominal because just two values), alive or dead; total sample size less than 1000: Fisher's exact test. Note that even though you only count the number of live flies, you started with 50 eggs, so you know how many flies died.
  24. One measurement variable, leaf length; one nominal variable, identity of tree: one-way anova
  25. One measurement variable, uvula length; one nominal variable, professional vs. church vs. non-singers: one-way anova
  26. Two nominal variables, magnet on or off, quadrant; total sample size over 1000: chi-square or G-test of independence
  27. One measurement variable, jump length; one nominal variable, baby or not: two-sample t-test or one-way anova
  28. One measurement variable, income; one ranked variable, dental condition: Spearman rank correlation
  29. One measurment variable, number of oranges; one nominal variable, tree type: one-way anova
  30. Beta (or power, which is 1-beta; don't say beta AND power); standard deviation of polar bear weight

Return to the Biological Statistics syllabus