The flashcards below were created by user
on FreezingBlue Flashcards.
What does a significant result on the Omnibus test imply?
- The null is false
- The degree to which it is false is not implied
What are omnibus tests?
Types of stats tests that tell us whether or not the explained variance in data is more significantly than the unexplained variance
What is the error rate per comparison (PC)?
The probability of making a type 1 error at any given comparison
What is the familywise error rate (FW)?
- The set of comparisons that are made is called a family according to Tukey
- This is the probability that at least one comparison from a family will contain a type 1 error
What is the error rate for FW?
1-(1- the error rate per comparison) x the number of comparisons
How can we control the FW rates?
- Use a more stringent error rate per comparison (such as.025)
- Divide the desired overall significance level by the number of comparisons you plan to make
- Assess the probability associated with each comparison with more stringent value
- This is known as the Bonferroni Method
What is the alternate way of performing the Bonferroni method?
- Multiply the associated probabilities by the number of comparisons
- Compare this to the .05 significance level
For FW, what is the significance rate?
For factorial analysis of variance what are the IV known as?
For factorial analysis of variance what are the experimental conditions known as?
The levels of the factor
How are factor levels compared?
- Each factor level is compared with every level of the other factors
- This ensures that the design contains every combination of the factorial levels
What is a 3x2x2 design?
There are three factors, one with three levels and two with two
What is the main effect?
The effect of an independent variable on a dependent variable averaging across the levels of any other independent variables
What else do factorials allow us to analyse?
The interaction between variables