ANOVA
- Created by: kaytlhone101
- Created on: 06-05-18 21:01
View mindmap
- ANOVA
- When
is one-way ANOVA used
- Three or more conditions
- Measured on interval or ratio scale
- Datapoints are normally distributed
- One-way ANOVA analyses the differences between three (or more) means of the DV (while accounting for the spread of the data)
- Appropriate
hypotheses
- If one-tailed, the hypothesis should ALSO indicate direction of differences among the groups
- Refers to the effect of the IV on the DV.
- Homogeneity
of variance
- If Levene’s is non-significant THEN pick your usual post-hoc procedure (e.g., Tukey, listed under “equal variances assumed)
- IF Levene’s is significant THEN pick a post-hoc procedure from the “equal variances not assumed” section
- Effect
size
- Partial eta squared (ranges from 0 to 1)
- ANOVA
variance
- Total variance
- Variance between different conditions/groups due to manipulation of the IV (numerator)
- Variance among different participants due to uncontrollable sources (error) (denominator)
- Total variance
- Sphericity
- If Mauchly’s is non-significant THEN read the top row on SPSS (“Sphericity assumed”) for the omnibus results
- If Mauchly’s is significant THEN read one of the other three rows on SPSS for the omnibus results
- Reporting
results
- (F (?, ?) = ?, p = ?, ?p2 = ?)
- What type of ANOVA
- Report all post-hoc results and direction of differences
- When
is one-way ANOVA used
Comments
No comments have yet been made