In: Statistics and Probability
Hypothesis Testing and Confidence Intervals
The Reliable Housewares store manager wants to learn more about the purchasing behavior of its
"credit" customers. In fact, he is speculating about four specific cases shown below (a) through (d) and
wants you to help him test their accuracy.
b. The true population proportion of credit customers who live in an urban area exceeds 55%
i. Using the dataset provided in Files perform the hypothesis test for each of the above speculations (a) through (d) in order to see if there is an statistical evidence to support the manager’s belief. In each case,
oUse the
Seven Elements of a Test of Hypothesis, in Section 7.1 of your textbook (on or about Page 361) or the Six Steps of Hypothesis Testing I have identified in the addendum.
oUse α=2%for all your analyses,
oExplain your conclusion in simple terms,
oIndicate which hypothesis is the“claim”,
o Compute the p-value,
o Interpret your results,
ii.Follow your work in (i) with computing a 98% confidence interval for each of the variables
described in (a) though (d). Interpret these intervals.
iii.
Write an executive summary for the Reliable Housewares store manager about your analysis,
distilling down the results in a way that would be understandable to someone who does not
know statistics. Clear explanations and interpretations are critical.
Location | Income ($1000) |
Size | Years | Credit Balance ($) |
Rural | 30 | 2 | 12 | 3,159 |
Rural | 31 | 2 | 4 | 1,864 |
Rural | 37 | 1 | 20 | 2,731 |
Rural | 27 | 1 | 19 | 2,477 |
Rural | 33 | 2 | 12 | 2,514 |
Rural | 44 | 1 | 7 | 2,995 |
Rural | 42 | 2 | 19 | 3,020 |
Rural | 30 | 1 | 14 | 2,583 |
Rural | 50 | 2 | 11 | 3,605 |
Rural | 35 | 1 | 11 | 3,121 |
Rural | 27 | 2 | 1 | 2,921 |
Rural | 30 | 2 | 14 | 3,067 |
Rural | 22 | 4 | 16 | 3,074 |
Rural | 53 | 1 | 7 | 2845 |
Suburban | 32 | 4 | 17 | 5,100 |
Suburban | 50 | 5 | 14 | 4,742 |
Suburban | 66 | 4 | 10 | 4,764 |
Suburban | 63 | 4 | 13 | 4,965 |
Suburban | 62 | 6 | 13 | 5,678 |
Suburban | 55 | 7 | 15 | 5,301 |
Suburban | 54 | 6 | 14 | 5,573 |
Suburban | 67 | 4 | 13 | 5,037 |
Suburban | 22 | 3 | 18 | 3,899 |
Suburban | 39 | 2 | 18 | 2,972 |
Suburban | 54 | 3 | 9 | 3,730 |
Suburban | 23 | 6 | 18 | 4,127 |
Suburban | 61 | 2 | 14 | 4,273 |
Suburban | 46 | 5 | 13 | 4,820 |
Suburban | 66 | 4 | 20 | 5,149 |
Suburban | 74 | 7 | 12 | 5394 |
Suburban | 66 | 7 | 14 | 5036 |
Urban | 54 | 3 | 12 | 4,016 |
Urban | 55 | 2 | 9 | 4,070 |
Urban | 40 | 2 | 7 | 3,348 |
Urban | 51 | 3 | 16 | 4,110 |
Urban | 25 | 3 | 11 | 4,208 |
Urban | 48 | 4 | 16 | 4,219 |
Urban | 65 | 3 | 12 | 4,214 |
Urban | 55 | 6 | 15 | 4,412 |
Urban | 21 | 2 | 18 | 2,448 |
Urban | 37 | 5 | 5 | 4,171 |
Urban | 21 | 3 | 16 | 3,623 |
Urban | 41 | 7 | 18 | 4,828 |
Urban | 48 | 2 | 8 | 3,866 |
Urban | 34 | 5 | 5 | 3,586 |
Urban | 67 | 5 | 1 | 5,345 |
Urban | 55 | 6 | 10 | 5,370 |
Urban | 52 | 2 | 11 | 3,890 |
Urban | 62 | 3 | 2 | 4,705 |
Urban | 64 | 2 | 6 | 4,157 |
Urban | 29 | 4 | 4 | 3,890 |
Urban | 39 | 4 | 15 | 4,183 |
Urban | 26 | 7 | 17 | 4,603 |
Urban | 44 | 6 | 5 | 3962 |
Urban | 25 | 3 | 15 | 3442 |
Here , the true population proportion = 0.55...
A new sample data is there...
Where , the sample population proportion = (Total no. of credit
customers who lives in urban area ) / Total no. of customers
= 24/ 55 = 0.436363636...
Now, to test :
H0 : The proportion of customers who lives in urban area = 0.55
vs
H1: The proportion of customers who lives in urban area > 0.55...
So, test statistic = (0.436363636 - 0.55) / SQRT ( 0.436363636 * (1- 0.436363636) / 55 )..........
= -1.699318926.....
P-VALUE = prob( test statistic > -1.699318926)
= 0.044629554......
now, α=2%...so, p-Value > 0.02...
So, we can't reject the null hypothesis in favor of the alternative
hypothesis
So, here, we can't say that the proportion of customers who lives
in urban area exceeds 55%..
So. the result is, the proportion of credit customers who lives in
urban area does not exceed 55%..while using our willingness to make
a Type I error α = 0.02..
confidence interval = 0.436363636 ± ( z* SQRT (
0.436363636 * (1- 0.436363636) / 55 )) , where z = p-value when
α=2%...
so,z = -2.053748911...
so, confidence interval = ( 0.299025933 , 0.573701339).........
so, there is a 98% chance that the true proportion of credit customers who lives in urban area is in between (0.299025933 , 0.573701339).........