In: Statistics and Probability
The following summary statistics represent the hours of battery
life for 4 brands of laptops. At the 0.05 significance level, do
the data provide sufficient evidence to conclude that a difference
exists between the the battery life of the fourdifferent
brands?
Column n Mean Std. Dev. Brand_A 8 7.69 1.62 Brand_B 8 10.44 2.45
Brand_C 10 8.55 2.52 Brand_D 8 5.91 1.30
here null hypothesis: Ho :there is no difference exist among mean of battery life of the four different brands
alternate hypothesis:Ha: at least 2 batteries have different mean life,
Applying ANOVA on above data:
Group | ni | x̅i | S2i | ni*(Xi-Xgrand)2 | (ni-1)*S2i |
A | 8 | 7.69 | 2.624 | 1.852 | 18.37 |
B | 8 | 10.44 | 6.003 | 41.180 | 42.02 |
C | 10 | 8.55 | 6.350 | 1.435 | 57.15 |
D | 8 | 5.91 | 1.690 | 40.903 | 11.83 |
grand mean= | 8.1712 | 85.371 | 129.37 | ||
SSTr | SSE | ||||
Source | SS | df | MS | F | |
between | 85.37 | 3.000 | 28.4571 | 6.60 | |
within | 129.37 | 30.000 | 4.3124 | ||
total | 214.74 | 33.0000 | |||
here test statistic F=6.60
at 0.05 level and (3,30) degree of freedom ; critical value F =2.92
as test statisitic is higher than critical value ; therefore we reject null hypothesis
we have sufficient evidence at 0.05 level to conclude that at least 2 batteries have different mean life,