Biological Data Analysis: Exam 1 Answers
- Beer amount is a nominal variable, because there are just two values; number of flies is a measurement variable. The ID of each trap is not a variable, because you only have one observation of one variable for each trap.
- Soil type is a nominal variable; number of apples is a measurement variable
- Two nominal variables (soapy vs. not soapy, elementary vs. college) and small sample size so Fisher's exact test
- One nominal variable (right vs. left) and small sample size so exact test of goodness-of-fit
- One nominal variable (the time period of each scratch, morning, day, or evening) and small sample size, so exact test of goodness-of-fit.
- Two nominal variables (taurine vs. no taurine, symptom type) and small sample, so Fisher's exact test
- Perfect vs. imperfect is a nominal variable, expression amount is a measurement variable
- Two nominal variables (MUNK17 vs. control, live vs. dead) and small sample, so Fisher's exact test
- Coffee amount: measurement (there are 8 different values), number of tomatoes: measurement, tomato variety: nominal
- Two nominal variables, temperature and genotype, and small numbers, so Fisher's exact test
- Activity type is a nominal variable, time of day is a measurement variable
- Three nominal variables: city pair, hot vs. cool, and long vs. short hair, so Cochran-Mantel-Haenszel test
- This is the probability of getting the observed result (55% left thumb on top), or a more extreme result, by chance if the null hypothesis were true
- The null hypothesis is that the proportion of white geese is the same at the two locations.
- Size is a ranked variable; tastiness is a measurement variable
- You should use a smaller P-value as your critical value, because it is very unlikely that singing to warts will make them go away; therefore you expect that most P-values less than 0.05 would be false positives
- You need the effect size, the difference in left-handed proportion between the children of smokers and non-smokers that you hope to see. You got points off for saying you needed beta, because you're given the power of 90 percent and thus know the beta is 10 percent; beta is not the "additional information" the question asks for.
- There is one nominal variable, right vs. left, and I specified in class that the sample size was small, so exact test of goodness-of-fit
- The probability that the first child does nothave cystic fibrosis is 3/4; the probability that the second does not have CF is 3/4; so the probablity that both the first and the second do not have CF is 3/4 x 3/4=9/16.
- There are two nominal variables, year and injury type, and a small sample size, so Fisher's exact test
Return to the
Biological Data Analysis syllabus
Return to John McDonald's home page
This page was last revised September 25, 2012. Its URL is