Class Notes (1,034,033)
CA (592,564)
UVic (3,064)
PSYC (418)
Lecture 19

# PSYC 201 Lecture Notes - Lecture 19: Random AssignmentPremium

2 pages38 viewsFall 2016

Department
Psychology
Course Code
PSYC 201
Professor
Stephen Lindsay
Lecture
19

This preview shows half of the first page. to view the full 2 pages of the document.
PSYC 201 Dr.Lindsay Lecture 19: Factorial Designs
2x2 between subjects allows you to examine the differences between the four conditions
and the differences that they each hold
o Can characterize the results in various ways
i.e. the mean being represented for each group then to be compared to each
other, then added across the two (marginal mean) conditions
o Interested in several different questions
Was there a main effect of one variable?
Overall were there a different score for two conditions?
Was there an interaction between the conditions?
o Is the overall averages going to be different from each other?
Line graphs can show interaction and main effects more than bar graphs
Need to average the subgroups of each conditions to determine if there is a main effect to
show whether or not there is a significant difference between the two conditions
A good strategy to determine if there is a difference; if the group sizes are about the same
then the mean of the groups would be roughly in the middle of the measures (see lecture
slides) [will be on the exam]
o Determine if the two points are distant form each other/if there is a significant
difference
You can tell at a glance whether or not there will be a tendency of an interaction, because
the lines are not parallel
Whether or not it is statistically significant depends on how large the groups are, how
much variability there are in the groups, and how big the interaction is.
The effect of one IV depends upon the other IV
Tendency of interaction is consistent of group congruency between two variables
Interaction effect: "cross-over" interaction, the direction of difference is important, there
can be a flip of the direction of the effect
Simple Effect Tests (post-hoc comparisons)
When a significant interaction occurs you want to perform a simple effect test as a way of
further exploring the data and looking at the effect of one IV separately at each level of
conditions
Includes doing two t-tests to see the difference between the conditions at each level
Orthogonally crossed= independently crossed
2x2 factorial design has three effects
Factorial Designs with Subject (Participant) Variables [non-manipulated]
Such as gender, ethnicity, age
Cannot change these variables but can create a factorial experiment using these variables
as the comparison groups
o i.e. boys vs. girls
Unlikely to happen by a fluke of chance
2x2 Within-subject Factorial Design
Same people tested at the time of each condition
Same individual tested in all four conditions
Would have advantages; not a large group of people, flukes of random assignment
couldn't account for your assignment because everyone has to go through each condition
Would try to counterbalance the order of each task for each individual
2x2 mixed-model factorial design
One of the IV are manipulated between subject and one IV is manipulated within subjects
2x3 Factorial Design
More than two levels of one or the other IV's
find more resources at oneclass.com
find more resources at oneclass.com