EPIB301 Exam Questions with Correct Answers
One-way ANOVA
comparing means along 3 or more groups
ANOVA assumptions
- Independence
-Normality (normally distributed)
-Homogeneity (underlying variances of populations of each group are equal)
Bartlett's test
checking equal variances (homogeneity) - k> 2 groups
When to use Bartlett's test
-before ANOVA
- requires normality within groups
If you fail to reject Ho using Bartlett's test...
it is OK to proceed with ANOVA
ANOVA Steps
1. Set hypotheses
2. set error rates
3. calculate test statistic
4. determine significance
5. interpret findings
, ANOVA null hypothesis
Ho: M1 = M2=...Mn
ANOVA alternative hypothesis
Ha: at least one inequality among M
Grand mean
the overall average of all data points across all groups combined, calculated by summing all
observations and dividing by the total number of sample
Sum of Squares
this is the raw, total amount of "spread" in your data - calculate it by squaring the distance
between data points and the mean, then adding them all up.
Mean square
This is the is the SS adjusted for the size of your study. You take that "Total Variation" (SS) and
divide it by its degrees of freedom
k
number of group
n
number of total observations
F statistic can only take on BLACK values, and the F distribution is skewed...
positive, to the right
One-way ANOVA
comparing means along 3 or more groups
ANOVA assumptions
- Independence
-Normality (normally distributed)
-Homogeneity (underlying variances of populations of each group are equal)
Bartlett's test
checking equal variances (homogeneity) - k> 2 groups
When to use Bartlett's test
-before ANOVA
- requires normality within groups
If you fail to reject Ho using Bartlett's test...
it is OK to proceed with ANOVA
ANOVA Steps
1. Set hypotheses
2. set error rates
3. calculate test statistic
4. determine significance
5. interpret findings
, ANOVA null hypothesis
Ho: M1 = M2=...Mn
ANOVA alternative hypothesis
Ha: at least one inequality among M
Grand mean
the overall average of all data points across all groups combined, calculated by summing all
observations and dividing by the total number of sample
Sum of Squares
this is the raw, total amount of "spread" in your data - calculate it by squaring the distance
between data points and the mean, then adding them all up.
Mean square
This is the is the SS adjusted for the size of your study. You take that "Total Variation" (SS) and
divide it by its degrees of freedom
k
number of group
n
number of total observations
F statistic can only take on BLACK values, and the F distribution is skewed...
positive, to the right
