In: Statistics and Probability
2. A company claims that a certain vitamin contains 100% of your recommended daily allowance of vitamin C. You suspect that they are not putting as much vitamin C as they claim in their vitamins. The standard deviation of percent of daily allowance of vitamin C is 5%. You take an SRS of 5000 vitamins and find the average percent of daily allowance to be 99.8%.
(a) Use the appropriate methods to determine if your fears are true.
(b) Estimate the mean percent of your daily allowance of vitamin C in these vitamins.
(c) Was part (a) statistically significant? Why or why not?
(d) Was part (a) practically significant? Why or why not?
(e) You want to create a confidence interval with a confidence level of 90% and a margin of error no more than 0.7%. What is the smallest sample size you can use and get these requirements?
Answer:
a)
Given,
Null hypothesis Ho : u = 100
Alternative hypothesis Ha : u < 100
alpha = 5% = 0.05
Standard deviation s = 5
Sample size n = 5000
Sample mean = xbar = 99.8
Degree of freedom = n - 1
= 5000 - 1
= 4999
Standard error SE = s/sqrt(n)
substitute values
= 5/sqrt(5000)
= 0.0707
consider,
test statistic z = (xbar - u)/SE
substitute values
= (99.8 - 100)/0.0707
= -2.82885
z = - 2.83
Corresponding p value = 0.002
Here we observe that, p value is less than alpha, so we reject Ho.
b)
Estimation of the mean percent = 99.8%
c)
Yes, we can clearly say that the part a was statistically significant due to p < 0.1, so we reject Ho.
And the claim isn't true.
d)
Yes, it is practically significant.
It is due to that the some percent of vitamins are lost by product handling.
e)
E = 0.7
At 90% CI, z value = 1.65
Sample size n = (z*s/E)^2
substitute values
= (1.65*5/0.7)^2
= 138.9031
Sample size n = 139