Lecture 6.Experimental Design

Variable pretest condition experimental group control

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: sign Independent Variable Pretest Condition Experimental Group Control Group No Pretest Pretest     If administration of the pretest has no effect on the DV the posttest scores will be the same for the people in the two control groups and for the people in the two experimental groups If administration of the pretest has an effect on the DV the posttest scores of the participants in the two control groups will be different and the posttest scores of the participants in the two experimental groups will be different (we may then just use the posttest only groups’ data)   Disadvantage ◦  Order effects – the order of the conditions may have an effect on the dependent variable   Practice effects – an improvement in performance due to practice on the task in previous conditions   Fatigue effects – a deterioration in performance due to fatigue, boredom or distraction resulting from previous conditions   Contrast effects – a change in responses due to a contrast with a previous condition 5 5/20/13     Changing the time interval between the conditions Counterbalancing ◦  Complete Counterbalancing! Condition A Condition B Condition B Condition A ◦  Randomized Counterbalancing ◦  Partial Counterbalancing (Latin Squares) – each condition appears at each ordinal position and each condition precedes and follows each condition one time Order 1: ABDC Order 2: BCAD Order 3: CDBA Order 4: DACB   Advantage ◦  Ensures groups are equivalent (at least on the matched variable) while eliminating concerns with order effects   DIsadvantages ◦  Costly and time consuming ◦  Participants can rarely be perfectly matched ◦  Matching may serve as a demand characteristic   In order for a study using an independent groups design to be considered a true experiment random assignment must be used   In order for a study using a repeated measures design to be considered a true experiment counterbalancing must be used 6...
View Full Document

This note was uploaded on 01/23/2014 for the course PSYC 217 taught by Professor Victoria during the Winter '11 term at UBC.

Ask a homework question - tutors are online