Collecting data from people in a sample or population, often involves asking them questions and recording their responses. One of the most common survey methods is the questionnaire, where participants answer a set of questions on a printed or online form. Other survey methods, like personal interviews, can be used to collect data in a similar way.
The quality of the data collected depends a lot on the quality of the questions being asked. Some important considerations when designing questionnaires are:
The two main types of questions used in questionnaires are either 'closed' or 'open'.
Our goal in surveying is to obtain accurate information about from our sample that we can make useful inferences about the population as a whole. As we saw previously, inaccurate data can arise from sampling errors and/or measurement errors.
For surveys a common form of measurement error is response bias. This is when participants respond inaccurately or falsely to questions. This can occur for a wide variety of reasons:
Non-response to part or all of a survey can also lead to biased results. For example, posting a survey on the location of a nuclear waste site will likely only see those with a strong opinion or directly impacted make the effort to answer. This would bias the results as we are excluding those with moderate views. Subjects may choose not respond if:
In summary, a well designed survey is key to obtaining quality data. Further detail on some aspects of survey design can be found in our investigation.
Which of these questions are fair?
Do you feel that the TV news is an inaccurate portrayal of life’s problems?
Don't you think this newspaper is biased?
Do you prefer the look and feel of thick lush carpeting in your living room?
Do you take these extra strength multi-vitamins to supplement your diet?
None of them
Adults attending a local cinema were asked the following question:
“How many times did you see a movie at this cinema last year?”
Is this an example of Sampling Error or Measurement Error?
Sampling error
Measurement error
What type of Measurement Error is this?
The scale provided is inadequate.
This is an assumption based question.
Poor and/or leading question wording.
What is the main reason why this question is poor?
Poor and/or leading question wording
The people being surveyed need more information.
Relies too heavily on respondent memory.