A bookstore wants to determine how many books the people in the surrounding neighborhood read per month on average. They survey each customer who enters their store for one week. Identify any bias in this method. If appropriate, suggest a method more likely to produce a random sample.

Question

anonymous · Answer

@Fellowroot are you writing

fellowroot · Answer

I think I wrote too much and it kicked me off

anonymous · Answer

wow thank you @Fellowroot ! i appreciate it

fellowroot · Answer

The bookstore wants to know how many people in the SURROUNDING NEIGHBORHOOD read per month on average yet they survey each customer who enters their store for one week. The bias is that they are prejudging that each customer who enters and that they survey is indeed from their surrounding neighborhood.

 A better method that is more likely to produce a random sample is to ask each person before they fill out the survey if they live in the surrounding neighborhood. This way the bookstore can actually collect the information they are interested in.

 Another bias that should be considered is that they are prejudging that the customers will answer honestly and accurately. Just because a person is handed a survey does not guarantee that the person will fill it out correctly or in such a way so that it reflects the truth. A person can misinterpret a question or brag about the number of books they have read. 

Also another bias is that the bookstore is prejudging that whoever is charge of looking at the data from the surveys is doing an accurate job. Many times people are overlooking surveys as their job and these people can make mistakes. ( I personally worked for a company where this one girl had a job counting tick marks on surveys, she always tried to hide her work and get out of her job by hiding the surveys under her other paper work. FIY)

fellowroot · Answer

Another problem is that the bookstore's survey only takes place during one week. Doing this surely wouldn't accurately reflect the people who read book in the neighborhood since not everyone is going to be stopping by the book store during some random week. The survey should take place over a longer time interval to better represent the community.

fellowroot · Answer

Even another problem is that the bookstore doesn't seem to take into consideration if a single person stops by more than once during the survey time period. If a single person fills out a survey more than once then the data collected will not be accurate. 

And the last biggie!

The fundamental problem with their survey is that there are going to be plenty of people who live in the surrounding neighborhood who may not go to this bookstore at all and therefore are not represented. The book store could be a newly opened shop that isn't well known so almost no one shows up.

anonymous · Answer

Thanks @Fellowroot