Sec. 5.3 Survey Design

The two most important factors that affect the validity of a survey are; The sample chosen for the population that is being surveyed The design of the questionnaire used to collect the data.

The three steps in designing a questionnaire are; 1.Identify the population that you want to gather information on 2.identify the purpose of the survey 3.design the questionnaire itself

We identify the population because it allows us to then select a representative sample and to direct our questions towards that population. We need to know the purpose of the survey so we can ask questions that will give us information relative to this purpose.

Open vs. Closed Questions When designing your survey, you have to decide whether the question will be open, closed or a combination of the two types. An open question is answered by the respondent in his or her own words. Ie: What should students do about the tuition rise in post secondary education. This can be answered in many ways and opinions can be used.

Open vs. Closed Questions Continued…… A closed question would be answered by checking a box or circling one of the given responses. The respondents are given choices that they must select from. What advantages/disadvantages are there to each of these types? Respondents prefer shorter surveys to longer ones.

Pointers Keep questions clear and concise. Wordy or complex questions can confuse or turn off respondents. If content of a survey is controversial or of a sensitive nature, be sure to check your question to ensure the respondents can answer them as comfortably as possible. This may require that the survey be confidential or that the respondent remains anonymous. This may also lead to more truthful answers.

Pointers Avoid use of technical language or acronyms. Acronyms should be expanded unless target audience are already familiar with them.

Pointers Avoid questions that ask for opinions about two different things because often they cause confusion among respondents and produce results that cannot be interpreted. Eg. If the respondent answers the question “Do you believe that air bags are unsafe and expensive?” with an answer of NO, you may conclude that they believe that airbags are safe and affordable when they really think that air bags are safe yet are still expensive.

Pointers Avoid biased questions. A question is biased if the wording is such that it leads to a particular conclusion. Eg. The question “Given the failure of welfare in the United States, do you feel welfare programs should be eliminated?” This leading question will most likely result in agreement by your respondents.

Homework p. 335 # 1-7, 9, 11 or 12