Chapter 3 – Measurements, Mistakes, and Misunderstandings
When reading the rsults of a study it is important to understand how the information was collected and
what was asked. Many answers to questions can be changed just by the wording. Example: “buy” vs.
Deliberate bias is when questions are worded to get a certain answer. “Do you agree that…” is a big one.
Appropriate wording should not indicate what response is right.
Unintentional bias is when the wording of a question is misinterpreted by a large group of people.
Specificity and clarification are key to proper questions.
Desire to please is when the participant gives an answer they think is the “right” one to the surveyor.
Participants will understate socially unacceptable habits and opinions, or the opposite based on who
they’re talking to.
Asking the uninformed is when people talk like they know stuff because they don’t want to admit they
don’t. Many people will state they know about something when it doesn’t even exist.
Unnecessary complexity is when the question is complex and people don’t know how to answer them.
Double negatives are a big one. Another example would be to ask 2 questions in one.
Ordering of questions can change results because one question may influence an answer to a future
Confidentiality (surveyor knows everything about participants, won’t reveal anything) and anonymity
(surveyor knows nothing about participant) affect things. Confidentiality is used when there are follow-
ups involved and anonymity can’t happen. Anonymity is used during questions on personal matters.
Open questions are when respondents are allowed to answer in their own words. Closed questions are
when they are given a list of alternative answers. Open questions can lead to results that are not easily
summarized. Some concepts are hard to define precisely. IQ tests are constantly changing because we
can’t agree on what specifically should be asked on them. Measuring attitudes and emotions are tough
as well. They are usually measured with closed questions.
Categorical variable can be placed in a category, but may not have logical ordering. If the categories
have a natural ordering they are ordinal variables. Example: strong