|
Step 5: Word all questions clearly and
unambiguously (I)
so that respondents understand what you are trying to
learn from them.
|
 |
|
Avoid these mistakes:
Double-Barreling: Avoid soliciting opinions about two
different issues in the same question, which forces the respondent
to impart one opinion that may not apply to both.
Example: Asking instructors, "How satisfied are
you with your salary and with the amount of prep-time you get
during the school day?"
Commentary: The question addresses both salary and prep time,
which should be addressed in separate questions.
Rewrite into two questions:
- How satisfied are you with your salary?
- How satisfied are you with the amount of prep-time you get
during the school day?
False Premising: Avoid embedding assumptions or opinions
in a question.
Example: Asking college administrators, "What practices
have you employed to get more faculty to use technology in their
classes?"
Commentary: The question assumes that all the administrators
want their faculty to use technology, when that may not be
true.
Rewrite into two questions:
- Have you taken an active role in encouraging your faculty
to use technology?
- If yes, what practices have you employed to bring that end
about?
Ambiguous Wording: Avoid vague, ambiguous words or phrases
that respondents need to interpret. The less clear your question
is, the less certain you can be about the meaning of their
responses.
Example: Asking instructors,"How many projects
did you assign to your students over the past six months?"
Commentary: Using the word "project" will yield ambiguous
results because there will be considerable variation in the
ways the instructors define it. Either the word "project"
must be clearly defined or the question needs to be rewritten
as a series of questions for different project types.
Rewrite: Circle how many of the following project types you
assigned to your students over the past six months.
| Reports |
0 |
1 |
2 |
3 |
4 |
5 |
| Experiments |
0 |
1 |
2 |
3 |
4 |
5 |
| Presentations |
0 |
1 |
2 |
3 |
4 |
5 |
| Information Models |
0 |
1 |
2 |
3 |
4 |
5 |
Overlapping Alternatives: Avoid combining multiple questions
into one question.
Example: Asking students, "Do you like using calculators
or do you like using spreadsheets?"
Commentary: It will be ambiguous which of the alternatives
respondents are addressing in their answers.
There are two ways this question could be rewritten:
- "Which do you like using more: calculators or spreadsheets?"
(if you want a preference expressed)
- As two questions: "Do you like using spreadsheets?"
and "Do you like using calculators?" (if you want
independent judgments of each)
Double Negatives: Avoid questions containing double negative
concepts that would be more clearly expressed as a positive concept.
Example: Asking community members, "Do you oppose
a law that would prohibit dress codes in public schools?"
Commentary: Many respondents would have to figure out the positive
version of this question before they could answer it.
Rewrite: "Do you favor a law that permits dress codes
in public schools?"
Intention to Act: Avoid seeking respondents' opinions
about hypothetical situations, unless you are really asking them
the question to learn something about them, and care less about
their opinions.
The following question asks teachers to venture an opinion
about some new software that they are about to be introduced
to in a workshop. At this point, they know nothing about it.
Question: "Will you use the software in your classes?"
The following is an example of a purpose that WOULD justify
the posing of the question before the workshop:
Let us assume that the teachers are required by their principal
to use the software, and the purpose of the question is to
see how many of them will follow the requirement, regardless
of whether they end up liking the software or not.
The following is an example of a purpose that would NOT
justify the posing of the question before the workshop:
Let us assume that teacher use of the software is voluntary and the evaluators
are trying to use the question to get a count of how many
will use it.
|
