In: Statistics and Probability
A biologist observes a sample of 25 cells to have a mean dividing time of 30 minutes with a standard deviation of 3.2 minutes.
a. Construct a 99% confidence interval for population mean dividing time.
b. A colleague claims that the mean dividing time is 32 minutes, but the biologist believes that it is less than 32 minutes. Carry out a hypothesis test and decide who is right - the biologist or the colleague.
c. State your conclusion in context of the hypothesis test and confidence interval.
d. What is the possible error made? What does the error mean in this problem situation?
a)
Level of Significance , α =
0.01
degree of freedom= DF=n-1= 24
't value=' tα/2= 2.797 [Excel
formula =t.inv(α/2,df) ]
Standard Error , SE = s/√n = 3.2/√25=
0.6400
margin of error , E=t*SE = 2.7969
* 0.6400 = 1.7900
confidence interval is
Interval Lower Limit = x̅ - E = 30.00
- 1.790041 = 28.2100
Interval Upper Limit = x̅ + E = 30.00
- 1.790041 = 31.7900
99% confidence interval is (
28.21 < µ < 31.79
)
b)
Ho : µ = 32
Ha : µ < 32 (Left tail
test)
Level of Significance , α =
0.010
sample std dev , s =
3.2000
Sample Size , n = 25
Sample Mean, x̅ =
30.0000
degree of freedom= DF=n-1=
24
Standard Error , SE = s/√n = 3.2/√25=
0.6400
t-test statistic= (x̅ - µ )/SE =
(30-32)/0.64= -3.125
critical t value, t* =
-2.492 [Excel formula =t.inv(α/no. of tails,df) ]
p-Value = 0.0023 [Excel formula
=t.dist(t-stat,df) ]
Decision: p-value<α, Reject null hypothesis
Conclusion: There is enough evidence that biologist is correct
c)
There is enough evidence to reject the null hypothesis that true mean is 32 minutes
there is 99% confidence that true ,mean lies within confidence interval
d)
Type I error could be made
Type I error is concluding that mean is less than 32 but infact it is 32 minutes