In: Statistics and Probability
"sleep latency time in minutes"
CBT |
melatonin |
control |
7 |
12 |
27 |
9 |
11 |
31 |
5 |
14 |
32 |
9 |
12 |
22 |
8 |
11 |
37 |
7 |
13 |
24 |
9 |
14 |
39 |
10 |
13 |
42 |
9 |
14 |
25 |
6 |
14 |
39 |
Use the data to compute the following:
1. One-way ANOVA
2. Tukey’s HSD
3. Eta squared post hoc test
4. Effect size
5. T obs
6. F obs
for soltution i used Excel.
part1)
SUMMARY |
||||
Groups |
Count |
Sum |
Average |
Variance |
CBT |
10 |
79 |
7.9 |
2.544444 |
melatonin |
10 |
128 |
12.8 |
1.511111 |
control |
10 |
318 |
31.8 |
51.28889 |
ANOVA |
|||||||
Source of Variation |
SS |
df |
MS |
F |
P-value |
F crit |
|
Between Groups |
3187.4 |
2 |
1593.7 |
86.38807 |
1.85E-12 |
3.354131 |
|
Within Groups |
498.1 |
27 |
18.44815 |
||||
Total |
3685.5 |
29 |
we reject H0 based on pvale and critical statistics. there is significant difference
part2) Tukey HSD
SE=sqrt(0.5*MS_w*(1/n_i+1/n_j))
q = abs(differece)/SE
Difference |
n(group1) |
n(group2) |
SE |
q |
||
CBT |
melatonin |
-4.9 |
10 |
10 |
1.35824 |
3.607611 |
control |
CBT |
23.9 |
10 |
10 |
1.35824 |
17.59631 |
melatonin |
control |
-19 |
10 |
10 |
1.35824 |
13.9887 |
tukey statistics n=30 and group=3 is 3.49
so that we can infere that all the pairs are significantly differe.
3) Eta squared post hoc test
neta = ss.between/ss.total
= 3187.4/3685.5 =0.864849
4) Effect size = 0.864849 which means the independent variable explained the variation of dependent varible by 86.48 percent. which is large effect size.
5) T statistics = sqrt(F) =sqrt(86.38807) =9.294519
6) F statistics =86.38807 = MSST/MSE