Question

Dr. Walter Wendler, President of WTAMU, is trying to decide if and where the University should build a parking garage and gives you the task to evaluate parking needs for WT’s campus. He asks you to distribute a survey that he has created and hands you a survey about campus parking that asks questions such as how often do you use campus parking, which hours do you need parking, what locations is more parking needed, and general parking views. How would you go about distributing the survey? Which sampling method did you choose and why would that be most appropriate? What are the pros and cons of this method? Discuss at least 1 other sampling method you could use to collect your data and explain why you would NOT use this/these. In your answer, be sure in include selection bias, non-response error, sampling error, and measurement error.

Answer 1

Survey is a sampling technique in which we make use of a simple random probability sampling method.

Simple random probability samples have a characteristics of the equal chance of getting selected out of the entire target population with the known probability of the target population who is interested to use the parking garage.

Surveys with these questionaire will help us to make an unbaised inferences about the target population.

In this example, Target population is the students and staffs who uses bicycles, motor cycles, and cars, etc. I will prefer to distribute the surveys to 30 randomly selected people who come in their vehicle to the campus in the morning hours from 9 to 10am for the five different days of a week.

Instead of simple random sampling, I would prefer to use a stratified sampling distribution method. Because the number of students and staffs ratio is different here. I will group the staffs and students separely, and distribute the survey to 30 randomly selceted staffs and students on an alternate days in a week. The advantage of the grouping will give more facus on the target population of interest who would need the parking garage. It would give a specific focus to the students and staffs population's requirements.

Selection bias will arise in a probability random sample, since i do not consider the true proportion of the population who need parking garage. It has a chance of visiting guests who uses the parking garage not considered at the time of sampling. If a set of students who come in campus for doing part time students only attend the classes in the evening time, i will miss them to include in the population of interest. In every survey, the errors are due to the random probability sampling methods. They are the coverage error, is due to omitting or not including all the parties involved in the need of parking garage. Non-response error is due to the method does not provide any incentive or rewards for the students or staffs, or the population found it difficult to repond due to the time constraint in the morning hours. It will end up in getting a poor sample data that may not reflect the interest of the target population. When it is a simple random probability sampling method, we wont be following the random process that generates the random numbers which is the perfect way to improve the reliability of the servey data. and the the true proportion of the population getting selected is not known at the time of surveying. This will create a sampling error. Measurement error happens when we dont consider the variability of the sample data from person who is distributing the surveys. For example if the surveys are distributed by the male or female, or staffs, or students, will get a different numbers due to the method of measurement that has built-variation.