6. Test yourself on power and sample size
1.The main purpose of a power analysis is to ensure that you get statistically significant results
2. Sampling variation is a common cause of type I errors
False. The main purpose is to estimate sample size for an experiment or power or effect size if sample size is fixed
True. Type I errors or false positive results arise from sampling variation.
3. Large sample sizes help to reduce the chance of type II errors
4. If the null hypothesis is true but we reject it, that is a type II error
True. and a power analysis helps us to estimate the correct number of replicates
False. It is a Type I error or false positive result
5.In order to detect an effect size of one standard deviation we need about 17-22 subjects, with an 80-90% power, a 5% significance level and a two-sided test.
6. If we increase the significance level from 0.05 to 0.10 we need more subjects.
True. It is useful to remember that to detect an effect size of one SD we need about 20 animals/group
False. We need fewer animals, other things being equal
7. The resource equation method is based on the law of diminishing returns
8. In a power analysis reducing the standard deviation increases the significance level
True. Beyond a certain point adding an additional subject does little to increase power
False. Significance level is specified. Reducing the SD increases the probability of detecting a statistically significant difference. But that is not the question
9. p-values show the probability that a difference as great as or greater than that observed in the experiment could have arisen by chance sampling variation
True. Well done.