 |
| Picture from mines.edu |
Everyone’s favorite subject J:10 questions for which it might be useful to know the answers…
1. What is a reasonably accurate method of estimating sample
size for an experiment? Which of these would you need to do this estimation?
An estimate
of variability
Power
p-value
useful effect size
2. What is the difference between the following
observational studies?
Cohort
Cross
sectional
Case control
3. Which of the study types in Q2 would best describe this
study…
In a colony consisting
of rhesus and pigtail macaques, look at the historical records and sort them into diarrhea
and no-diarrhea groups. Determine the
proportion of rhesus and pigtails in each group.
4. What is the
dependent variable in Q3? (hint –
another term for dependent variable is outcome measure)
5. True or false: if statistical evaluation of your study
results in you rejecting the null
hypothesis, you can deduce that there is
a meaningful effect of your treatment.
6. True or false: a high (ie larger #) p value reduces the chance of false
negatives but increases the chance of false positives.
7. You calculate
the confidence interval (CI) for an experiment comparing the means of 2
samples using a p value of 0.05:
- The CI does not span zero and
- The 2 values of the confidence interval are far apart
How would you interpret these 2 findings ?
8. Which of these are considered to be continuous data?
- a. degree of liver fibrosis
- b. liver enzyme values
- c. fish species
- d. # of dogs in each US State
- e. number of male students in a classroom
9. Which of these statistical tests would generally be used
on continuous data?
- a. ANOVA
- b. Shapiro-Wilk Normality test
- c. Students’ t test
- d. Mann Whitney
- e. Kruskal Wallis
10. If you are comparing the means of 4 samples with
one dependent variable (data are continuous), which statistical test are you most
likely to use?
Answers on the right sidebar: Q #12
Answers on the right sidebar: Q #12
No comments:
Post a Comment