Types of Bias (College Board AP® Statistics)
Study Guide
Written by: Mark Curtis
Reviewed by: Dan Finlay
Types of bias
What is a biased sample?
A sample is biased if the method used to create the sample consistently under-represents or favors certain groups in the population
Biased samples are, on average, unlikely to be representative of the patterns and behaviors of the population
Though it is still possible to select a biased sample that is coincidently representative of the population!
Biased samples can overestimate or underestimate the property of the population that you are studying
Which types of bias do I need to know?
You need to know the following different types of bias
Undercoverage bias means some of the population you are trying to sample either has a lower chance of responding or are fully excluded for some reason
e.g. calling all home owners to ask about local wi-fi speed will miss responses from those who rent
Voluntary response bias means only those who volunteer will be heard
e.g. for a call-in show, or an optional survey
Voluntary responses can often contain strong opinions!
Nonresponse bias means some of those selected for the sample never give a response
e.g. they are impossible to contact or purposefully do not reply
Response bias is a noticeable trend in the responses that suggest they are not answering a question fairly or truthfully
The problem is often the wording of the questions (question wording bias) that either confuse individuals or lead them towards a certain answer
e.g. "Isn't it a waste of time to ...?"
e.g. "Do you prefer cats over dogs?" (versus "dogs over cats?")
Other problems include individuals
being embarrassed of their answers
trying to please the interviewer
not understanding the question
being asked to self-report (write about themselves)
Testing questions first can reduce response bias
Worked Example
A town mayor wants to ban alcohol in a public park. She selects 50 residents from the four roads surrounding the park and pressures them for a response to the question:
"Don't you agree that it is about time we banned alcohol in our beautiful park? [Yes / No]"
She obtains answers from 45 of the selected residents but 5 of the selected residents refuse to answer the question.
Describe the different types of bias that are present in this situation.
Answer:
The mayor is likely to have undercoverage bias as only the four roads surrounding the park were sampled, meaning no other residents in the town had their responses recorded
There is nonresponse bias as 5 residents, who were asked the question, refused to answer it
There is response bias as the wording of the question being asked is leading them to believe that it should be banned (question wording bias)
Further response bias is likely as some residents may be too embarrassed to say no, or may want to please the mayor who is asking the question
Note that voluntary response bias has not been included as residents were pressured to give an answer (it did not appear to be very optional or voluntary)
Last updated:
You've read 0 of your 5 free study guides this week
Sign up now. It’s free!
Did this page help you?
