Mind map
Independent One-Way ANOVA

Mind map: independent one-way ANOVA
Map content
Independent One-Way ANOVA
The Need for ANOVA
- One-Way ANOVA
- 1 IV with more than 2 levels
- Estimates population means under different levels of the IV are different
- Independent one-way ANOVA - between-subjects
- Repeated measures one-way ANOVA - within-subjects
- Extension of the t-test
- Why not t-tests?
- When p < .05 - prepared to reject the null
- Still a 5% chance of Type I error
- 3 IV levels = 3 t-tests
- Each of these t-tests - 5% chance of making a Type I error
- Overall Type I error rate would be higher
- Familywise error rate - probability that at least one of a family of comparisons will result in Type I error
- Provides corrected significance level
- a = 1-(1-a)^c
- a - error rate for family of comparisons
- (1-a) - level of confidence were working to for each comparison
- c - number of comparisons
- a = 1-(1-a)^c
- 3 IV levels
- 3 comparisons - afw = .143
- Omnibus tests control this
- Provides corrected significance level
- One-Way ANOVA
F-ratio
- Null hypothesis - there is no difference
- F = variance between IV levels/variance within IV levels
- Close to 0 - small variance
- Further from 0 - large variance
Variance
- Between IV levels
- Manipulation of IV (treatment effects)
- Individual differences
- Experimental error
- Random error
- Constant error
- Within IV levels
- Individual differences
- Experimental error
- Random error
- Partitioning the variance
- 1) calculate the means for each IV level
- 2) calculate the grand mean - sum of IV level means, divided by number of IV levels
- 3) calculate within IV levels variance - sum of squared differences between individual values and IV level mean
- 4) calculate between IV levels variance - sum of squared differences between each IV level and the grand mean
- Between IV levels
Assumptions
- Normality
- DV should be normally distributed under each level of the IV
- Homogeneity of variance
- Variance in the DV should be equivalent
- Levene's test checks
- Significant result = violation
- Welch's F corrects
- Equivalent sample size
- Sample size under each level should be roughly equal
- Independent of observations
- Scores under each level should be independent
- Non-parametric equivalent - Kruskal Wallis test
- Normality
Degrees of freedom
- Difference between the number of measurements made and number of parameters estimated
- Calculate df for estimates of
- Between IV level variance
- df = k-1
- Within IV level variance
- df = N - k
- Between IV level variance
Post-hoc tests
- Secondary analyses used to assess which IV level mean pairs differ
- Only when F is significant
- Run as t-test but include correction
- Tukey Honestly Significant Difference (HSD)
Effect size
- Partial eta^2
- How much variance in the DV is explained by manipulation of the IV
- SSm/SSm+SSr
- Cohen's d
- Magnitude of the difference between IV level means
- d = x1-x2/(sd1+sd2/2)
- Partial eta^2
Information
- Description
- Psychology Mind Map on Independent One-Way ANOVA, created by Isabel LONGDEN on 16/01/2021.
- Created by
- Isabel LONGDEN
- Views
- 2
- Rating