In: Statistics and Probability
come up with a study (something you are interested in learning more about) that would require the use of an ANOVA and calculation of the f-statistic. What I'm lookin' for here is that you would be able to identify a study that would require this statistical procedure vs. the procedures we've learned previously (t and z).
Walk us through the steps in the process of calculating the f-stat
Dear student, please comment in the case of any doubt and I would love to clarify it. Here, I have taken the example and tried to show the step involved in ANOVA test.
Using the following summary data, perform a one-way analysis of variance using α=.01.
Var1=[10.45]**2=109.2
Var2=[12.76]**2=162.82
Var3=[11.47]**2=131.56
MSerror=(109.2+162.82+131.56) / 3= 134.53
Calculating the remaining error (or within) terms for the ANOVA table:
df(error) = 90−3 = 87
SS(error) =(134.53)(90−3) =11703.82
Intermediate steps in calculating the variance of the sample means:
Grand mean = (50.26+45.32+53.67) / 3=49.75
Sum of squares (SS(means)) = 35.25
Var(means) = 35.253−1 = 17.62
MS(between) = (17.62)(30) = 528.75
Calculating the remaining between (or group) terms of the ANOVA table:
df(groups) = 3−1 = 2
SS(group) = (528.75)(3−1) = 1057.5
Test statistic and critical value
F = 528.75 / 134.53 = 3.93
F(critical)(2,87) = 4.86
Decision: fail to reject H0