One-Way ANOVA

Lecture 3

What does ANOVA stand for?

Analysis of Variance

Why would you use a One-Way ANOVA?

to compare results amongst 3 or more groups

How many IVs and DVs are there in a One-Way ANOVA?

1 IV and 1 DV

What would the null hypothesis be in a One-Way ANOVA?

“the averages from all groups are the same”

What should you consider when performing a One-Way ANOVA?

• means

• variances

What are the important assumptions for ANOVA?

• assume normal distribution (normality)

• the groups are independent of each other

• homogeneity of variance (distributions have equal/similar variance, Levene’s test was used)

• DV in continuous or “scale” measurement (interval or ratio)

Will an ANOVA test tell you which of the groups is most successful at xyz?

No; it will only tell you if there is a difference between groups

Which Post-Hoc tests should be performed when the ANOVA is found to be significant?

• Bonferroni

• Tukey’s honestly significant difference

• Scheffe’s

• Fisher’s Least Significant Difference

Why do you perform Post-Hoc tests after significant ANOVA results?

Post-Hoc tests are meaningless if the main ANOVA is not significant

What are the precautions for conducting multiple ANOVA tests?

• multiple tests = multiple DVs

• probability of making a Type I error increases

What should be included in the results portion of the analysis?

• only tested facts

• p value

• mean

• SD

• sometimes F statistics