In: Statistics and Probability
Question: Thirty years ago, the Big Three automakers dominated car sales in this country. Backthen, Ford accounted for 28% of sales, GM for 26%, and Chrysler for 16%. Foreign companies accounted for 30% of sales. Data for 158 recent car purchases was collected to see if there is a significant shift in this pattern:
Car Maker |
Ford |
GM |
Chrysler |
Foreign |
|
Cars Sold |
42 |
25 |
20 |
71 |
a)Use the Goodness of Fit Hypothesis Test to find if the pattern from 30 years ago is still in place, using the 0.05 level of significance.Show all steps of the hypothesis test: Hypotheses, Alpha, Distribution, Sketch with Critical Value, Calculate Test Statistic by hand and put on Sketch, Decision, Conclusion.
b)Would your conclusion change if you changed your level of significance level to 1%? Show why or why not.
a)
null hypothesis:Ho: pattern of car purchase follows which was in place 30 years ago
alternate hypothesis: pattern of car purchase differs from which was in place 30 years ago
alpha=0.05
distribution -chi square:
for (categories-1=3) df and 0.05 level critical value =7.815
applying chi square goodness of fit test:
observed | Expected | Chi square | |||
category | Probability(p) | Oi | Ei=total*p | R2i=(Oi-Ei)2/Ei | |
Ford | 0.280 | 42.000 | 44.240 | 0.113 | |
GM | 0.260 | 25.000 | 41.080 | 6.294 | |
Chrysler | 0.160 | 20.000 | 25.280 | 1.103 | |
Foreign | 0.300 | 71.000 | 47.400 | 11.750 | |
total | 1.000 | 158 | 158 | 19.261 |
from above chi squae test statistic =19.261
as test statistic falls in rejection region we rject null hypothesis
we have sufficient evidence at 0.05 level to conclude that pattern of car purchase differs from which was in place 30 years ago
b)
for 0.01 level and 3 df crtiical value =11.345
as test statisitc is higher than critical value out conclusion will remain same at 0.01 level