In: Statistics and Probability
A random sample of five observations from three normally distributed populations produced the following data: (You may find it useful to reference the F table.)
Treatments | ||||||||||
A | B | C | ||||||||
24 | 18 | 31 | ||||||||
26 | 21 | 27 | ||||||||
19 | 27 | 21 | ||||||||
24 | 23 | 16 | ||||||||
30 | 19 | 30 | ||||||||
x−Ax−A |
= |
24.6 | x−Bx−B | = | 21.6 | x−Cx−C | = | 25.0 | ||
s2AsA2 | = | 15.8 | s2BsB2 | = | 12.8 | s2CsC2 | = | 40.5 | ||
Click here for the Excel Data File
a. Calculate the grand mean. (Round intermediate calculations to at least 4 decimal places and final answer to 4 decimal places.)
b. Calculate SSTR and MSTR. (Round intermediate calculations to at least 4 decimal places and final answers to 4 decimal places.)
c. Calculate SSE and MSE. (Round intermediate calculations to at least 4 decimal places and final answers to 4 decimal places.)
d. Specify the competing hypotheses in order to determine whether some differences exist between the population means.
H0: μA = μB = μC; HA: Not all population means are equal.
H0: μA ≤ μB ≤ μC; HA: Not all population means are equal.
H0: μA ≥ μB ≥ μC; HA: Not all population means are equal.
e-1. Calculate the value of the F(df1, df2) test statistic. (Round intermediate calculations to at least 4 decimal places and final answer to 3 decimal places.)
e-2. Find the p-value.
p-value < 0.01
f. At the 1% significance level, what is the conclusion to the test?
Do not reject H0 since the p-value is not less than significance level.
Reject H0 since the p-value is less than significance level.
Reject H0 since the p-value is not less than significance level.
Do not reject H0 since the p-value is less than significance level.
g. Interpret the results at αα = 0.01.
We conclude that some means differ.
We cannot conclude that some means differ.
We cannot conclude that all means differ.
We cannot conclude that population mean C is greater than population mean A.
Overall total = 356
Grand mean Y̅.. = 356 / 15 = 23.73
Y̅ .. is overall mean
y̅i . is treatment mean
SS total = ΣΣ(Yij - & Y̅..)2 = 310.92
SS treatment = ΣΣ(Yij - & y̅i.)2 = 34.5333
MS treatment = ΣΣ(Yij - & y̅i.)2 / a - 1 = 17.2667
SS error = Σ(y̅i. - & Y̅..)2 = 276.4000
MS error = Σ(y̅i. - & Y̅..)2 / N - a = 23.0333
ANOVA table | |||||
Source | SS | df | MS | F | p-value |
Treatment | 34.53 | 2 | 17.267 | 0.75 | .4934 |
Error | 276.40 | 12 | 23.033 | ||
Total | 310.93 | 14 |
Part d)
H0: μA = μB = μC; HA: Not all population means are equal.
Part e)
Test Statistic :-
f = MS treatment / MS error = 0.7496
Part e-2)
P value = 0.4934
p-value > 0.10
Reject null hypothesis if P value < α = 0.01
Since P value = 0.49 > 0.01, hence we fail to reject the null
hypothesis
conclusion :- Treatment means are same
Do not reject H0 since the p-value is not less
than significance level.
We cannot conclude that all means differ.