Carry over effects in within subjects designs and their remedies.

Suppose a researcher is interested in whether their is any effect of the color of a maze on the speed with which a rat runs it. What are the dependent and independent variables?

Let's say the researcher creates five mazes. They have an identical number of blind alleys and turns in the correct path to the food reward, and each presents a different, but equally difficult (he believes) pattern. He paints each one a different color, (a) black, (b) white, (c) red, (d) green, and (e) blue. In the experiment, each of 12 rats will run each maze once and will be timed.

If the researcher suspects there might be each of these possible carry over effects, what is the simplest remedy?

• First describe the problem if the name is given, or name it if a description is given.
• Name the remedy and describe it.
• Write the order of the letter of the treatments.

A linear practice effect.

A linear fatigue effect.

A warm up effect (that is, a non-linear practice effect).

Any time the rat runs the black maze, he may be slower on the maze that follows.

(Answers - For best learning results: Don't peek until you try by yourself, first)

Confounds

According to the description of the experiment, there is a potential for a confound.

• What is the potential confound?
• What would be the easiest way to eliminate it?
• If it were eliminated the easiest way, what other problems would that create?
• What would be another way to eliminate the confound that would not have those problems, but that would be more difficult to implement.

(Answers)