Question

ANOVA

An experiment was conducted where 30 men were randomly assigned to a sugar pill (placebo), low, or high doses of Viagra. Their libido after taking the drug was measured on a 10 - point scale.The question is whether there is an impact on libido (rather than ED). The researchers consider an effect of interest to be a 2 - point difference in the response between doses on this scale. The variables are:

• Participants libido: the libido of the participant (taking the drug) on a 10 - point scale

• Dose: the dose level of Viagra administered to the participant: placebo = sugar pill or no medication, low, high

1. [Writing]. If the researchers wanted to have high power of detecting a 2 - point difference, what changes to their study design could they have made ? What would be the best change, statistically speaking? Explain.

2.What are the null and alternative hypotheses for testing the statistical question?

Answer 1

1) For the given study design we may increase the power of design to detect the difference of 2 point n scale incorporating many ways-

1. increase alpha (level of significance)
2. Conduct a one-tailed test (Probably this test is one-tailed because we increase the dose to get better outcome)
3. Increase the effect size (unless you’re coming up with a more precise way to measure your constructs, it’s likely that the effect size is a big as it’s going to get)
4. Decrease random error
5. Increase the sample size (The most important way to increase the power of design)
6. Replication: repeated measure for every subject.

2) The null hypothesis for the ANOVA can be given as

H0: The difference is not 2 point between the outcome of three doses

Alternate hypothesis:

H1: The difference is greater than 2 points on the scale between the three doses.