In: Statistics and Probability
Confidence Intervals
1. I want to estimate the average money spent on Xmas toys each week in December. I take a
sample of 40 individuals and determine that the mean spent is $210 with a standard
deviation of $32.00
a. For a 95% confidence interval, estimate the interval within which the true population
would fall.
b. If I reduced the interval size to just 90%, what would that interval be?
c. For the 90% confidence interval, calculate the maximum error of the estimate/mean?
2. The MBTA wants to estimate the average amount of money an individual spends at MBTA parking lots each week. A sample of 64 commuters was taken and the average amount paid was $44 dollars, with a standard deviation of $3.00.
a. What is the best estimated of the amount paid for parking?
b. Construct a 90% confidence interval of the population mean.
Answer)
1)
Given mean = 210
Standard deviation = 32
Sample size (n) = 40
As here population standard deviation is unknown, we will use t distribution to estimate the interval
A)
First we need to estimate the degrees of freedom, degrees of freedom is equal to n-1, 40-1, = 39
And confidence level is 95%
For degrees of freedom 39 and confidence level of 95%
Critical value t is 2.023
Now we need to find the margin of error
MOE = t*(standard deviation/√n)
= 2.023*(32/√40)
= 10.2356603304
Confidence interval is given by,
Mean - MOE, Mean + MOE
(199.764339669, 220.235660330)
We are 95% confident that true population mean lies in between 199.76 and 220.24
B)
In case of 90%, critical value is 1.685
Therefore,
MOE = 1.685*(32/√40)
= 8.52550057181
Confidence interval
= (mean - moe, mean + moe)
= (201.47449942819, 218.52550057181)
We are 90% confident that true population mean lies in between 201.47 and 218.53
C)
We know that error = t*(standard deviation/√n)
t for 90% confidence level and degrees of freedom 39 is 1.685
Standard deviation = 32 and n =40
Error = 1.685*(32/√40)
= 8.52550057181
2)
Answer)
A) best point estimate is the sample mean:
Which is $44
B)
As population standard deviation is unknown, therefore we will use t distribution
Degrees of freedom is equal to n-1, 64-1 = 63
And confidence level is 90
From t distribution critical value for 63 degrees of freedom and 90% confidence level is
= 1.669
Margin of error (MOE) = t*(standard deviation/√n)
MOE = 1.669*(3/√64)
= 0.625875
Confidence interval is given by
(Mean - MOE, mean + MOE)
(44-0.625875, 44+0.625875)
( 43.374125, 44.625875)
We are 90% confident that true population mean lies in between 43.37 and 44.63.