Question

An automobile manufacturer sold 30,000 new cars, one to each of 30,000 customers, in a certain year. The manufacturer was interested in investigating the proportion of the new cars that experienced a mechanical problem within the first 5,000 miles driven. (a)
A list of the names and addresses of all customers who bought the new cars is available. Describe a sampling plan that could be used to obtain a simple random sample of 1,000 customers from the list.

Each customer from a simple random sample of 1,000 customers who bought one of the new cars was asked whether they experienced any mechanical problems within the first 5,000 miles driven. Forty customers from the sample reported a problem. Of the 40 customers who reported a problem, 13 customers, or 32.5%, reported a problem specifically with the power door locks. (b)

Explain why 0.325 should not be used to estimate the population proportion of the 30,000 new cars sold that experienced a problem with the power door locks within the first 5,000 miles driven. (c)

Based on the results of the sample, give a point estimate of the number of new cars sold that experienced a problem with the power door locks within the first 5,000 miles driven.

Answer 1

A) In order to create a sampling plan, you need to follow the following 5 steps:
    1) Define the sample population: who are the customers you want to contact?
       The costumers who bought a new car on a certain year.
    2) Define the size of population: how many customers are you going to contact?
       Of the 30000 customers who bought a car, you want to contact 1000 customers.
    3) Define the contact options: how are you going to contact the customers?
       You have a list of names and addresses, therefore you can send a questionnaire via mail.
    4) Form a sampling frame: what is the time or contact frame to get in touch with your customers? 
       You will send the questionnaires and you will wait two months for the answer.
    5) Define the analysis method: is yours a qualitative or quantitative research?
     In your case, you want a quantitative research and therefore a probabilistic sampling.

B) The 32.5% probability refers to customers having issues with the power doors locks among the costumers who had problems, it does not consider the customers who did not have any problem or those who had problems after the first 5000 miles.

C) In order to find the probability of a power door lock problem if there have been problems within the first 5000 miles, we need to consider the whole sample:
P = 13 / 1000
   = 0.013

Therefore, 
N = 0.013 × 30000
   = 390

Hence, the number of new cars sold that experienced a problem with the power door locks within the first 5000 miles will be 390.