Chapter 7 - repeated measures design

What is repeated measures design?

Every subject is their own control since they have to participate in all conditions - both the control and experiment conditions.

What other terms can be used when talking about RMD?

Within-subjects design, inomgruppsdesign.

What are some advantages with RMD?

- it requires fewer participants than IGD (independent groups design)
- it is more convenient and efficient
- it is generally more sensitive than IGD
- error variation is generally less
- it is a good tool in seeing change over time in participants' behaviors

What is sensitivity?

The ability to detect the effect of the independent variable even if the effect is a small one

What are practice effects?

The changes that participants undergo with repeated testing

How do you eliminate practice effects?

By balancing them out with certain techniques

What designs can be used to eliminate practice effects?

Complete and incomplete designs (general term for these designs are counterbalancing).

What is a complete design, and what techniques are used to arrange the order of the conditions?

A design where the practice effects are balanced for each participant. By administering the conditions to each participant several times and using different orders each time, each participant can be considered a 'complete' experiment.

The order of the conditions can be arranged in one of two ways: **block randomization** and **ABBA counterbalancing.**

What is ABBA counterbalancing?

ABBA counterbalancing involves presenting the conditions in a random order followed by the opposite of that order (ABCCBA). An even number of repetitions is required with this technique.

What are the limitations with ABBA counterbalancing?

- it can only be used when practice effects are linear
- anticipation effects (when participants develops expectations about which condition occurs next in the sequence)

What is an incomplete design, and what techniques are used to arrange the order of the conditions?

A design where each condition is administered to each participant only once. It is essential that practice effects are balanced by varying the order in which the conditions are presented.

The techniques:
- All possible orders
- Selected orders

What is 'all possible orders' (incomplete design)?

The amount of orders equal to all the possible orders, which is N! (N factorial, where N are the conditions). For example, if there are two conditions (A B), then that is 2 factorial which is 2 orders (AB, BA). Five conditions \= 5 factorial \= 120 orders.

What is 'selected orders' (incomplete design)?

The number of selected orders will always be equal to some multiple of the number of conditions.

For example, if we have seven levels, we will need to select 7, 14, 21, 28, or some other multiple of seven orders to balance practice effects.

What are the two types of selected orders (incomplete design)?

Latin Square: each condition appears at each ordinal position once. Each condition also precedes and follows each other condition exactly once.

Random Starting Order with Rotation: you begin with a random order of the conditions and rotate this sequence systematically by moving each condition one position to the left each time. Like the Latin Square, each condition appears at each ordinal position but because of the rotation, the conditions always follows and precedes the same other conditions

What is differential transfer?

When performance in one condition differs depending on the condition that precedes it.

For example, the order in which you eat a cake will depend on how the tasting will go. If you eat a very sweet chocolate cake before eating a carrot cake, the chocolate taste will linger and affect the carrot cake condition.