Idk Flashcards
Factors that effect power
- Sample size
- Actually effect size
- Statistical significance
The main way to increase power is to increase sample size
One sample t test
Limited - don’t know standard
Two sample t test
More common to compare between two groups
Assumptions of a 2 sample t test
Data is normal
Homogeneity of variance
Samples are independent
Scale must be at interval/ratio
T tests
Allow us to compare two groups at a time
Analysis of variance (anova)
Allows us to compare the mean of several groups
-effect if multiple independent variables with 2 or more levels each
Potential types of variance in an anova
Assume that all of the populations have the same variance
Within v. Between
When the null is true, between = variance
When null is not true
Between > within
F ratio
Ratio of between to within is greater than 1
Post hoc comparisons
Attempt to control alpha and maintain it at .05
One sample t test
1 group compared to “known”
Two sample t test
IV has 2 levels (not related)
Paired sample
1 IV, 2 levels (related)
One-way anova
1 IV > 2 levels
Two or more variables are studied simulateously
Complex design
To describe a design in written form, include the name of the IV and the levels in parentheses
Writing designs
The overall effect of one independent variable
Main effect
What is one main effect possible for
Every independent variable in the design
The effect of one independent variable changes depending on the level of another IV
Interaction
Often used when the IV is a nominal variable
Bar graph
Provide easiest way to view interactions
Line graphs
