1/14
Looks like no tags are added yet.
Name | Mastery | Learn | Test | Matching | Spaced | Call with Kai |
|---|
No study sessions yet.
Steps of ANOVA
Hypotheses
Calculate overall mean, group means
Degrees of freedom between groups, within groups
Compute SSG: mean of squares among groups
Compute SSE: mean of squares within groups
Compute MSG and MSE
Calculate F stat
Use F table for critical value
Make decision
N
number of total observations of data (3 groups, 4 observations each, N = 12)
k
number of groups
DF between groups
k-1
DF among groups
N-k
SSG (mean among groups)
sum of n(group mean - total mean)². (n is the same for each group; number of groups does not change)
formula: sum of ni * (xi bar - x bar)²
SSE (mean within groups)
sum of (individual value - total mean)² for each group.
(ex: should have sum of 4 values per group if we have 3 groups of 4 observations)
formula: (x - x bar)²
MSG (mean of squares among groups)
SSG/(k-1)
MSE (mean of squares within groups)
SSE/(N-k)
F statistic
MSG/MSE
critical value
find f critical value using (k-1, N-k)
What do we do when we reject null hypothesis?
use fisher’s LSD to find out which groups are different
power of the test
correctly rejecting null hypothesis
what influences the power of the test?
increase sample size
increase alpha
decrease population sd
increase difference between hypothetical mean and population mean
which factors related to the power of the test can we control?
increase alpha
increase sample size