GRADUATE STATISTICS - MODULE 6 –
ANOVA EXAM 100% ACCURATE
LATEST UPDATED
Review: - ANSWERz-Test and One-Sample t-Test:
- Compare ONE sample man to a population mean
Independent Samples t-Test & Paired Samples t-Test:
- Compare TWO sample means against one another
Analysis of Variance (ANOVA) - ANSWER- Used to test mean differences between 2 or
more groups
- Provides the same result that a t-Test would
- Preferred for groups of >2, given the increased probability of making Type I errors,
based on alpha, with multiple t-tests (ie: if each test is run at a = .05, each test has a 5%
probability of making a type 1 error. For each test run, this is added to the % of error)
- The more t-tests run, the greater the probability of making errors (3 t-tests have a
14.26% probability of error)
- ANOVA controls for the problem of experiment-wise error rate b/c a single test is being
run
Experiment-wise error rate - ANSWEROverall probability of making an error in a study
- ANOVA controls for this because only 1 test is being done, so there is only 1
Experiment-wise error rate
Advantage of Variance (ANOVA) over t-test: - ANSWERCan be used to compare more
than 2 groups at once
In an ANOVA, whatever is used to control the groupings is known as a: -
ANSWERFactor
- The independent or quasi-independent variable that determines the groupings
- Independent variable that is manipulated
- Can also be a variable that is not manipulated
- More than 1 Factor can be present
- Within each factor are levels (k)
- Level refers to the # of groups that comprise a factor
Within each factor are: - ANSWERLevels (k)
Level refers to: - ANSWERNumber of groups that comprise a factor
, Analysis of Variance (ANOVA) Examples - ANSWERDetermines the groupings
Example 1:
- Funny movie
- Sad movie
- Neutral movie
= 1 Factor (movie types - example of a factor that can be manipulated), 3 levels (funny,
sad, neutral)
Example 2:
1 Factor (based on age - existing factor, not manipulated)
- Teenagers
- Young adults
- Middle aged
= 1 Factor (age), 3 levels (teens, YAs, and MAs)
Example 3: Factorial Design
2 Factors (each with 2 levels
Factor 1 (male/female)
- Men
- Women
Factor 2 (time of day)
- Morning
- Evening
In this case, men and women are 1 factor, because they each determine a grouping
Level (k) - ANSWER- Number of individual groups that make up a factor
Examples:
Funny movie, sad movie, neutral movie = 3 levels
Teenagers, young adults, middle-aged = 3 levels
Men, Women = 2 levels
Factorial Design - ANSWERStudy that has more than 1 factor
One-way ANOVA - ANSWERAn ANOVA test with a single Factor
Factorial ANOVAS are conducted for: - ANSWERFactorial Designs
Factorial ANOVAS are named for: - ANSWERthe number of factors they have, ie:
- 1 Factor: One-way ANOVA
- 2 Factors: Two-way ANOVA
ANOVA EXAM 100% ACCURATE
LATEST UPDATED
Review: - ANSWERz-Test and One-Sample t-Test:
- Compare ONE sample man to a population mean
Independent Samples t-Test & Paired Samples t-Test:
- Compare TWO sample means against one another
Analysis of Variance (ANOVA) - ANSWER- Used to test mean differences between 2 or
more groups
- Provides the same result that a t-Test would
- Preferred for groups of >2, given the increased probability of making Type I errors,
based on alpha, with multiple t-tests (ie: if each test is run at a = .05, each test has a 5%
probability of making a type 1 error. For each test run, this is added to the % of error)
- The more t-tests run, the greater the probability of making errors (3 t-tests have a
14.26% probability of error)
- ANOVA controls for the problem of experiment-wise error rate b/c a single test is being
run
Experiment-wise error rate - ANSWEROverall probability of making an error in a study
- ANOVA controls for this because only 1 test is being done, so there is only 1
Experiment-wise error rate
Advantage of Variance (ANOVA) over t-test: - ANSWERCan be used to compare more
than 2 groups at once
In an ANOVA, whatever is used to control the groupings is known as a: -
ANSWERFactor
- The independent or quasi-independent variable that determines the groupings
- Independent variable that is manipulated
- Can also be a variable that is not manipulated
- More than 1 Factor can be present
- Within each factor are levels (k)
- Level refers to the # of groups that comprise a factor
Within each factor are: - ANSWERLevels (k)
Level refers to: - ANSWERNumber of groups that comprise a factor
, Analysis of Variance (ANOVA) Examples - ANSWERDetermines the groupings
Example 1:
- Funny movie
- Sad movie
- Neutral movie
= 1 Factor (movie types - example of a factor that can be manipulated), 3 levels (funny,
sad, neutral)
Example 2:
1 Factor (based on age - existing factor, not manipulated)
- Teenagers
- Young adults
- Middle aged
= 1 Factor (age), 3 levels (teens, YAs, and MAs)
Example 3: Factorial Design
2 Factors (each with 2 levels
Factor 1 (male/female)
- Men
- Women
Factor 2 (time of day)
- Morning
- Evening
In this case, men and women are 1 factor, because they each determine a grouping
Level (k) - ANSWER- Number of individual groups that make up a factor
Examples:
Funny movie, sad movie, neutral movie = 3 levels
Teenagers, young adults, middle-aged = 3 levels
Men, Women = 2 levels
Factorial Design - ANSWERStudy that has more than 1 factor
One-way ANOVA - ANSWERAn ANOVA test with a single Factor
Factorial ANOVAS are conducted for: - ANSWERFactorial Designs
Factorial ANOVAS are named for: - ANSWERthe number of factors they have, ie:
- 1 Factor: One-way ANOVA
- 2 Factors: Two-way ANOVA
