Survey Research
What is a Survey?
A survey consists of asking many people the same questions.
Primary Advantage of using Surveys
1. Easy way to find out what people are thinking (very convenient and economical)
2. If a random sample is used, very easy to determine what everyone in country is thinking by only asking a few.
Primary Disadvantages of using Surveys
1. Can't make causal statements
2. Questionnaire construction is difficult
I Types of surveys
A Mail or Self-Administered Questionnaires
Advantages
wider and more representative sample
cost per response is low
easier for personal questions
no interviewer bias
Disadvantages
Must be self-explanatory
Lack control - not sure who actually filled it out
low response rate
B Telephone
Advantages
rapid data collection
interviewer can help explain questions
response rate is typically high
somewhat more control
Disadvantages
costs
reduced anonymity
potential interviewer bias
C Face to Face interviews
Advantages
very flexible (can probe and ask follow-up questions)
control is great (know who answers are actually coming from)
high response rate
Disadvantages
high costs
interviewer bias
Steps involved
1. Decide what variables you wish to obtain information about.
2. Decide what population is.
3. Decide the type of survey used (mail, phone, interview)
4. Write questions and decide how each response will be measured.
5. pre-test
6. collect data
7. analyze data.
Guidelines for Writing Questions
Simple rules of thumb:
1. Use very simple words and grammar.
Aim you questions at about an eighth grade reading level.
Avoid jargon and using words that people might not know what they mean.
For example:
What is your annual income?
How much money do you make each year?
2. Avoid leading questions.
That is, try to avoid swaying how people should answer the question.
Should try to make respondents feel that all answers or responses are ok and legitimate.
Extreme example:
"Right now as many as 6 MILLION illegal aliens have made their way through the U.S. borders and are living throughout America. Taxpayers are paying out BILLIONS in tax dollars for social security, welfare, unemployment, public housing, free health care, bilingual education and even legal defense costs for illegal aliens. An estimated 3 MILLION more illegals will sneak through the U.S. borders this year. The problem is becoming worse.
Do you personally approve of having your tax dollars spent to support illegal aliens?
3. Be care in the words you chose. Try to eliminate biased results in the words you choose. Choice of words can greatly affect the outcome.
Try to use neutral language. Do not bias people one way or the other.
However, more subtle ways this can be done.
For example: You get very different types of responses to these questions
Do you think women shouldn't be allowed to have abortions?
Do you think women should be forbidden to have abortions?
Do you think it should be against the law for women to have an abortion?
Do you think it should be a crime for a women to have an abortion?
Do you think women should go to jail for having an abortion?
Moreover, must avoid using words that people have strong emotional reactions to.
Try to use very neutral language when asking questions.
Also think about how you are presenting information when asking questions.
What are you highlighting in each sentence. (order matters - words same)
Do you support a woman's right to choose what she does with her body, when it comes to the issue of abortion?
When it comes to the issue of abortion, do you support a woman's right to choose what she does with her body?
4. Avoid double-barreled questions.
Don't ask two questions at once.
For example:
Do you think that affirmative action programs should be discontinued and the money saved spent on helping disadvantaged children learn how to read?
5. Avoid negative items.
Try to avoid using the word not in questions.
People often miss this word when reading a question
6. Decide if you should use a direct or indirect question.
Direct = just ask them what you want to know.
Indirect = Don't directly ask them what you want to know.
Appropriate when trying to find out about things that the respondent is unwilling or unable to reveal in direct terms. People like to provide socially appropriate answers = social desirability bias.
So, if you want to know if people have sexist attitudes towards women.
You probably don't come out and ask: Do you have sexist views towards women?
You might ask other more indirect questions.
Such as,
Do you think that women overstate the amount of discrimination they face?
Do you think that women's liberation movement has gone too far?
Do you think that women are better off now that they have the opportunity to work outside of the home?
When using indirect or even direct measures for that matter. It is better to use more than one question to try and get at the issue at hand.
Using multiple questions allows you to get at the issue from different angles. Typically, provides more valid and reliable information.