Chapter 7&8 Flashcards
- participants are in different conditions
- Can’t place people in their condition, it’s predetermined
- To show causation you have to create equivalent groups through random assignment
between subjects
- Requires fewer participants
- Sensitive to small differences between means
- Use counterbalancing
Within Subjects
T-Test: two or more conditions
- compares the means of two groups
- Assumptions
- Normal distribution, interval or ratio data
- Homogeneity of variance
Independent groups (sample)
the participants in conditions are not
connected to each other in any sort of way
When would independent groups (sample T test) be used
- Between subjects
- Random assignment
- Participant variables
Dependent groups (paired sample)
the participants are connected to
each other
When to use dependent groups (paired sample)
Between subjects
■ Matched groups design
■ Dyads
○ Within subjects
■ Manipulated variables
■ Repeated measures
Uses of single factor
Discover nonlinear (curvilinear) effects
■ Three levels of arousal (low, moderate, high)
○ Test specific alternative hypothesis
Interpolating:
you make estimates between points on a graph
Extrapolating:
you make estimates outside points on a graph
Multiple T-Tests:
increases likelihood of type I error
ANOVA:
comparing group means
IV is continuous:
line graph preferred, bar graph
acceptable
IV is discrete:
use a bar graph, line graph is inappropriate
factor
independent variable
