1. You are trying to determine what the best combination of diet and exercise provides the most weight loss. You decided to try three different diets (Atkins, Paleo, McDonalds only), along with no dieting, and two different forms of exercise (Yoga, high cardio), along with no exercise. You try each combination possible, and track weight loss for each. What type of Experimental Design is this?

Completely Randomized Design

Latin Square Design

Completely Randomized Block Design

Factorial Design

2. Consider an experiment where we wanted to determine which brand of soil increased plant height the most. We tested three different brands, and had 4 replications for each brand. Of the 12 observations, 3 plants just didn't grow. What type of bias, if any, would this be considered?

Nonresponse

Response

Sampling

No bias

3. Choose all that apply: Sample statistics

are always known

are estimated from parameters

are calculated from samples

require all known subjects of a population

4. Choose all that apply. For a cluster sample

All subjects have an equally likely chance of being selected

Clusters are usually determined by convenience

All subjects will be measured in a selected cluster

Clusters have to be of equal size

An equal number of subjects will be measured in every cluster

5. A response bias occurs when

The sample is not selected at random

Subjects were influenced to respond a certain way

Subject did not respond

The population was undercovered