Chapter 15-1: The Public Opinion Survey
Introduction
A survey is a series of questions designed to discover people's opinions, habits, beliefs, awareness, or knowledge. This page discusses how to prepare and distribute a survey.
Preparing for a survey
Goal: What's your goal for this survey? What do you want to discover?
Examples:
What are the working habits of full-time students?
How does age affect opinions about pregnancy outside of marriage?
Hypothesis: Make at least two guesses of the outcome of the survey.
Examples:
Most full-time students only have time to work part-time jobs.
Older full-time students tend to work more hours than younger full-time students.
Planning your questions
Prepare ten to fifteen questions for your survey. Don't ask essay-style, open-ended questions when you want to be able to easily analyze your survey results. Instead, provide survey takers with a selection of answers.
Question Types
Multiple Choice
How many hours do you work during the week?
None
1-20
21-40
More than 40 hours
What is your gender?
Male
Female
Other
Yes/No/I don't know
Tip: When surveying knowledge, always include an "I don't know" option. This prevents answers from being skewed.
How many countries are in the European Union?
28
18
I don't know
Rating Scale
Tip: It's important to use the same scale on multiple rating scale questions in order to avoid confusion and distorted answers.
On a scale of 1 to 5, rate your satisfaction with your school's class sizes.
1 - Very Dissatisfied
2 - Dissatisfied
3 - Neutral
4 - Satisfied
5 - Very Satisfied
Nominal Choices
Tip: Always include an "Other" option. This allows you to receive answers for items you didn't think of.
Circle all of the fruit juices you drink each week:
Orange
Apple
Grapefruit
Grape
Mango
Other(s): ______________________
Structuring questions
The beginning questions in the survey should distinguish the groups you're studying. For example, if you're studying how race or ethnicity affects opinions or habits, then your first question should determine race or ethnicity.
Provide clear questions and consistent answers. It's important that questions don't have answers that are too similar to one another. Use the tips above for Yes/No/I Don't Know questions or questions that use a rating scale.
Distributing a survey
Give clear instructions. At the beginning of the survey, write a short introduction for your survey explaining its purpose. Tell respondents how to complete the survey. See the sample survey below for an example.
Who will you survey? You may decide to survey a wide group of people in order to get a variety of opinions. Or you may want to select a narrower group so you can study the opinions or beliefs of that specific group.
How many people will you distribute the survey to? A larger sample will be more representative of the group at large, but sometimes you may have limited capabilities. If you're working with a group, each person in the group can survey a certain number of people.
How will you distribute the survey? You will also need to decide how to distribute your survey. Some options are:
Email
Social media
In person
By phone
By mail
Useful Words and Phrases
Satisfied / dissatisfied
Agree / disagree
Positive / negative
Neutral
In your opinion / In your view
Low / moderate / high
Slightly / somewhat / moderately
Very / extremely / strongly
Not at all
Refused
Choose all that apply
All of the above
Do you think...
Do you approve or disapprove of...
Do you consider yourself...
Do you view ________________________ as a problem?
How do you think others view ________________________________?
How important is _______________________?
Question words such as:
Does / Do / Did...
Are / Is...
How often...
How many...
What / Where / When / Who / Why / Which...
Should / Could / Would...
Sample Survey
The following is a model survey designed to gather information on public opinion regarding a particular topic. This survey demonstrates one specific format. Others exist. The natural sequence that follows the administration of an opinion survey is a data analysis and then a presentation, often in the form of a technical report.