In: Statistics and Probability
1. Does the mean cost of a 1999 Honda Civic differ by location? To answer this question, a researcher with a lot of free time collected data on recent sales of these cars for four major cities, which is shown in the table below.
Portland | Seattle | San Fran | Baltimore |
4998 | 5878 | 3977 | 3546 |
5341 | 5393 | 5105 | 4310 |
4875 | 4764 | 4779 | 4052 |
5127 | 4805 | 4006 | 3037 |
6861 | 6325 | 5248 | 3828 |
5745 | 5888 | 4000 | 4225 |
5793 | 6752 | 4462 | 4045 |
5043 | 5546 | 4346 | 3649 |
4738 | 5149 | 5104 | 3165 |
4276 | 5588 | 6306 | 4995 |
Notice: It’s 10 cars for each city. They’re not in any particular
order.
Calculations notes
You’re going to go through the calculations to find the p-value
from the data, but you can check all the calculations by using
Excel, a calculator, or something else.
The calculations for this module are way more complicated than
we’ve seen before, so please ask if you’re not sure. There’s also
an Excel file on D2L where you can see the formulas to see where
each number comes from.
a. Write hypotheses appropriate to test the question here.
b. A comparison boxplot of the prices is shown below. Based on the boxplot, does it appear the mean selling price for the cars is the same for every city? Explain your answer with evidence from the boxplot.
c1. Calculate the mean price for each city.
c2. Calculate the variance of the means.
c3. Multiply that value by 10 to get the MST. (10
because its 10 observations / city.)
d1. Calculate the variance in price for each
city.
d2. Calculate the mean of the variances. This is the
MSE.
e. Calculate the F-ratio as F = MST/MSE
f. To calculate the p-value, we need the numerator df
and the denominator df. What are those?
g. What is the p-value for the test?
h. Based on your p-value, make a conclusion with the
context of this situation.