Repeated Measures Designs Repeated Measures Type of design

Repeated Measures Designs

Repeated Measures • Type of design in which the same participants are used in every group and so are exposed to every level of the IV. • Participants must be measured on the DV multiple times. • Ensures that the groups are identical; so eliminates many E. V.

Quick Class Experiment • Imagine that you are being asked to rate scary media images for a study on children. Rate the pictures using the following scale • 1 = Not scary • 2 = A little Scary • 3 = Moderately Scary • 4 = Very Scary • 5 = Extremely Scary

Switch Groups

Order Effects • When the position of a treatment (level of IV) in a series determines, in part, the participants’ response. e. g. , you “leave some room” in the first rating because you do not know what is coming next.

Carryover Effect • The effects of one treatment persist or carryover and influence responding to the next treatment. e. g. , the first picture “anchors” your view of the second picture.

Are there possible order or carry-over effects?

What if we want to add another level?

How many sequences will we need? Do we have enough people to adequately cover each sequence?

Differential Carryover Effect • Response to a treatment depends on which specific treatment came before it. – An unforeseen and unavoidable problem! – Which alien sequence might produce this effect?

Counterbalancing Dependent Variables • Order effects aren’t just for IVs. • How you order questions in your survey can also show these effects. • As participants get bored or tired, order effects can show up in some questions. • Counterbalancing questions can even this out within your groups.

Question Order Discussion • Imagine your group wants to collect opinions on _______. • Write 3 questions that if ordered differently might produce order effects?

Types of Counterbalancing

Within Subjects Counterbalancing • Every participant gets all possible sequences. – ABBA design • Reasonable if you only have 2 levels. Coke Pepsi Coke But then they are responding multiple times to the same stimulus so other problems arise.

Within Groups Counterbalancing • Different participants get different sequences but each sequence is given to the same number of participants. • More common.

Complete Counterbalancing • All possible sequences are presented

Incomplete Counterbalancing • Random selection of sequences chosen.

Why bother with repeated measures? Eliminates a lot of extraneous variables related to participants? How many alien movies have you seen? Do you have phobias about aliens?

Moral of the Story! • Beware of repeated measures, particularly if you have more than 2 levels of the IV.