3.7 How to Experiment Well

3.7 How to Experiment Well

Learning Target

• Describe how to randomly assign treatments using slips of paper or technology.

Random Assignment

In an experiment, random assignment means that experimental units are assigned to treatments using a chance process.

Random Assignment Methods

• Using slips of paper: Label each paper with a letter or number. Shuffle and hand out one to each volunteer. For example, if you have 100 volunteers and you want two groups, label 50 with A and 50 with B.

• Using a number generator: Label the volunteers with a number, then randomly generate the amount needed (ignore repeats). For example, if you have 100 volunteers and you want two groups, after you assign each of them a number, generate 50 numbers. The remaining 50 will be in the other group.

Open Ended

Describe how to randomly assign 60 volunteers to 3 equal groups.

Learning Target

• Explain the purpose of random assignment in an experiment.

Random Assignment

• Random assignment should distribute the volunteers in roughly equal numbers to each group. In fact, the random assignment should create two groups that are roughly equivalent with respect to every variable that might affect the response.

• The purpose of random assignment is to create groups that are roughly equivalent at the beginning of an experiment.

Multiple Choice

The purpose of random assignment is to create different groups.

• True

• False

Learning Target

• Identify other sources of variability in an experiment and explain the benefits of keeping these variables the same for all experimental units.

Other sources of variability

• Although random assignment should create groups of experimental units that are roughly equivalent at the beginning of an experiment, there are other variables that might have an impact on the response variable.

• Depending on the experiment, some factors to consider might be temperature, amount, time, etc.

Other sources of variability

• One reason to keep other variables the same for each subject is to prevent confounding.

• The second reason we keep other variables the same is to reduce the variability in the response variable, making it easier to determine if one treatment is more effective than another. 

Multiple Choice

One reason to keep all other variables the same is to reduce variability.

• True

• False

Multiple Choice

One reason to keep all other variables the same is to encourage confounding.

• True

• False