6 - Analysis of Variation Oneway Flashcards
What is the difference between t tests and ANOVA?
- t tests = comparing the means of 2 groups
- ANOVA = comparing the means of > 2 groups
What is the problem w/ doing multiple t tests instead of ANOVA?
Multiple t tests increases the chance of type 1 error (that the test will say a result is significant when it in fact is not)
With ANOVA, what 2 things are we concerned with?
1) Variability within each group
2) Variability between the groups
What is the formula for F value?
F = variance BETWEEN sample means / variance among individuals WITHIN each sample
What is the difference between SD and variance?
Variance = SD^2
What is the difference between SSE and SSW?
- Same thing
- One is sum of squares error and one is sum of squares within groups
How do you calculate df?
of groups minus 1
What are you comparing for between group variance?
The means of each group
What are you comparing for within group variance?
How each variable in a group compares to the mean of that group
What are some assumptions for ANOVA one way?
- Simple random samples reduce bias
- We have independent samples, one from each of k populations
- For dependent measures (repeated measures on the same subject) a repeated measures ANOVA model is available
- The population from which the simple random samples are drawn must be normally distributed
- The sigma squared for each population must be equal (not significantly different, even when group means are different)
How do you know which critical F point to pick from the chart?
First number will be df for group 1, second number will be df for group 2
How must the calculated F value compare to the critical F value to be considered significantly different?
Calculated F value must be greater than critical F value
What does a significant F value tell you w/ ANOVA (when there are more than 2 groups)?
Tells you that at least 2 of the groups differ; doesn’t tell you which ones
What is the difference between SSB and SST?
- Same thing
- One is sum of squares between and the other is sum of squares treatment
For an example w/ 3 different treatment groups and N = 43 patients, what would be the df for between tx and within tx?
- Between tx = # of groups minus 1 = 3-1 = 2
- Within tx = # of patients minus # of groups = 43-3 = 3
When using the F distribution table, if a number you need isn’t on the table do you round up or down? Why
Round down so that it makes it that much harder to pass the test (being more conservative)
- So if you have 55, round down to 50 instead of up to 60
