Exam 2 Flashcards
What does ANOVA stand for?
Analysis Of VAriance
ANOVA description
compares the means of the samples by analyzing the variance between them
One-way ANOVA
- parametric
- only a single independent variable
- variances must be similar
- samples should be normally distributed
Kruskal Wallis ANOVA
- non-parametric
- most straight forward
Homogeneity definition
variances of all samples must be similar
What test is used to test homogeneity?
levene’s test
levene’s test
difference between highest and lowest variance
If you fail an assumption on normality only, what do you do?
- use correction on all samples
- use non-parametric test
If you fail an assumption on equal variances only, what do you do?
- use non-parametric test
If you fail an assumption on both normality & equal variance, what do you do?
Definitely use non-parametric test
What do these outputs tell us?
(one-way ANOVA: F=336.781, df=2.27,p<0.001)
(Kruskal-Wallis ANOVA: H=20.789, df=2, p<0.001)
- the means of three samples are not equal (at least 2 are different)
- treatment has “an effect” on hatch percentage, but what effect?
if guarding is important then…
A<BC
if guarding is NOT important, then…
AB<C
if care and guarding are both important….
A<B<C
Omnibus test
Looks at everything at the same time and provides single p-value to compare to alpha
- avoids alpha inflation
when should post hoc test be performed?
if omnibus result is significant
HSD stands for…
Honestly significant difference
Post hoc test for KW ANOVA
- more complicated
- Tukey’s HSD
What post hoc test is appropriate under most circumstances when using one-way ANOVA?
Tukey’s HSD
Post hoc test
test that is done after the omnibus test to help clarify the result
