Difference between revisions of "Make a Questionnaire"
(importing article from wikihow) |
m (Text replacement - "[[Category:R" to "[[Category: R") |
||
| (3 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
When a company, non-profit group, or politician needs to find out how their stakeholders or constituents feel, they often create and implement a questionnaire. The results can lead to re-branding, decision-making, and policy changes if the feedback is sound. Making a questionnaire can seem very straightforward, but unless it is designed properly, the results can be skewed and unreliable. | When a company, non-profit group, or politician needs to find out how their stakeholders or constituents feel, they often create and implement a questionnaire. The results can lead to re-branding, decision-making, and policy changes if the feedback is sound. Making a questionnaire can seem very straightforward, but unless it is designed properly, the results can be skewed and unreliable. | ||
| − | [[Category:Research and Review]] | + | [[Category: Research and Review]] |
== Steps== | == Steps== | ||
===Creating Questions=== | ===Creating Questions=== | ||
| − | #Decide what you want to learn from administering your questionnaire.<ref>http://www.surveysystem.com/sdesign.htm</ref> Ask yourself what data you require and how you will use it. This will help you come up with useful questions, as well as the order in which you will ask them. Ideally, the questionnaire will be short, so decide which of your goals are essential and which might be unnecessary. <ref>http://library.bcu.ac.uk/learner/writingguides/1.05.htm</ref> | + | #Decide what you want to learn from administering your questionnaire.<ref name="rf1">http://www.surveysystem.com/sdesign.htm</ref> Ask yourself what data you require and how you will use it. This will help you come up with useful questions, as well as the order in which you will ask them. Ideally, the questionnaire will be short, so decide which of your goals are essential and which might be unnecessary. <ref name="rf2">http://library.bcu.ac.uk/learner/writingguides/1.05.htm</ref> |
#Plan questions that will help you get the information you need. Begin with a broad span of questions, then narrow them down until each one relates to your goals in some way. Keep questions and answers simple, using as few words as possible. You may want to rely on open-ended questions, closed-ended questions, or a mix of the two. | #Plan questions that will help you get the information you need. Begin with a broad span of questions, then narrow them down until each one relates to your goals in some way. Keep questions and answers simple, using as few words as possible. You may want to rely on open-ended questions, closed-ended questions, or a mix of the two. | ||
| − | #Use closed-ended questions to gather specific answers. Closed-ended questions have a specific range of options respondents can choose from. <ref>http://fluidsurveys.com/university/comparing-closed-ended-and-open-ended-questions/</ref> These questions may be yes-or-no questions, true-or-false questions, or questions that ask the respondent to agree or disagree with a statement. Closed-ended questions may look like open-ended questions, but will have only a few options respondents can use to answer. Closed-ended questions might look like these: | + | #Use closed-ended questions to gather specific answers. Closed-ended questions have a specific range of options respondents can choose from. <ref name="rf3">http://fluidsurveys.com/university/comparing-closed-ended-and-open-ended-questions/</ref> These questions may be yes-or-no questions, true-or-false questions, or questions that ask the respondent to agree or disagree with a statement. Closed-ended questions may look like open-ended questions, but will have only a few options respondents can use to answer. Closed-ended questions might look like these: |
#* "Have you shopped here before?" | #* "Have you shopped here before?" | ||
#* "If so, how often do you shop here?" (This question would have a few explicit answers from which respondents could choose--"once a week" to "once a month," for instance) | #* "If so, how often do you shop here?" (This question would have a few explicit answers from which respondents could choose--"once a week" to "once a month," for instance) | ||
| Line 17: | Line 17: | ||
#* Open-ended questions are good for clarifying a previous answer--"Why do you feel this way?" | #* Open-ended questions are good for clarifying a previous answer--"Why do you feel this way?" | ||
#Ask questions in such a way as to avoid confusion and bias. Avoid leading questions especially; Leading questions indicate that the asker is looking for a certain answer and will limit what answers your responders are comfortable supplying. Either tailor possible answers or change the wording of your question so as to avoid leading your respondents to answer in a particular way. | #Ask questions in such a way as to avoid confusion and bias. Avoid leading questions especially; Leading questions indicate that the asker is looking for a certain answer and will limit what answers your responders are comfortable supplying. Either tailor possible answers or change the wording of your question so as to avoid leading your respondents to answer in a particular way. | ||
| − | #* You may consider asking the same question in different ways, which may reduce overall respondent bias and give you a better chance of finding the person's true opinion on a given topic.<ref | + | #* You may consider asking the same question in different ways, which may reduce overall respondent bias and give you a better chance of finding the person's true opinion on a given topic.<ref name="rf1" /> |
#* Questions should be worded so as to maximize clarity. Confused respondents will skew your data, so questions should be as understandable as possible. Avoid double negatives, unnecessary clauses, or unclear subject-object relationships. | #* Questions should be worded so as to maximize clarity. Confused respondents will skew your data, so questions should be as understandable as possible. Avoid double negatives, unnecessary clauses, or unclear subject-object relationships. | ||
| Line 26: | Line 26: | ||
#* Surveys delivered on the computer, in person, and by mail can utilize pictures, whereas phone interviews cannot. | #* Surveys delivered on the computer, in person, and by mail can utilize pictures, whereas phone interviews cannot. | ||
#* Respondents may be too shy to answer certain questions in person or by phone. Decide if you want to give clarifications to your questions if the respondent doesn't understand something; only interviews given by a live person can deliver clarifications. | #* Respondents may be too shy to answer certain questions in person or by phone. Decide if you want to give clarifications to your questions if the respondent doesn't understand something; only interviews given by a live person can deliver clarifications. | ||
| − | #* A computer survey will require the respondent to have access to a computer. If your questionnaire concerns private issues, a computer survey may work best. <ref>http://www.edelman.com/post/six-steps-creating-effective-questionnaire/</ref> | + | #* A computer survey will require the respondent to have access to a computer. If your questionnaire concerns private issues, a computer survey may work best. <ref name="rf4">http://www.edelman.com/post/six-steps-creating-effective-questionnaire/</ref> |
#Consider the order of your questions. The form of your questionnaire matters just as much as its content. You should aim to order questions so that they follow each other logically or mark clear shifts from section to section. Other kinds of questions may impact how a respondent fills out the questionnaire. | #Consider the order of your questions. The form of your questionnaire matters just as much as its content. You should aim to order questions so that they follow each other logically or mark clear shifts from section to section. Other kinds of questions may impact how a respondent fills out the questionnaire. | ||
#* You may want to order the questions so that if a person says yes or no to a certain question, they bypass any questions that don't apply to them. This will help keep the questionnaire focused and take less time to complete. | #* You may want to order the questions so that if a person says yes or no to a certain question, they bypass any questions that don't apply to them. This will help keep the questionnaire focused and take less time to complete. | ||
| Line 45: | Line 45: | ||
#Review your open-ended questions. See if your open-ended questions are working the way you need them to work. They may be too open, in which case respondents may ramble. They may not be open enough, in which case the data you received won't be as valuable. Ask yourself what role your open-ended questions are playing in your questionnaire and tailor them as needed. | #Review your open-ended questions. See if your open-ended questions are working the way you need them to work. They may be too open, in which case respondents may ramble. They may not be open enough, in which case the data you received won't be as valuable. Ask yourself what role your open-ended questions are playing in your questionnaire and tailor them as needed. | ||
#* As above, broad questions such as, "How do you feel while shopping here?" may not give your respondents enough direction. You could instead ask, "Would you recommend this store to your friends? Why or why not?" | #* As above, broad questions such as, "How do you feel while shopping here?" may not give your respondents enough direction. You could instead ask, "Would you recommend this store to your friends? Why or why not?" | ||
| − | #Decide how you will respond to missing data. Not all respondents will answer all questions, which may or may not be a problem for you. <ref>http://www.ats.ucla.edu/stat/seminars/survey_monkey.htm</ref> Ask yourself which questions are being skipped or answered incompletely, if at all. This may be due to the order of the questions, the wording of the questions, or the subject matter of the questions. If missing data is important, consider rewording skipped questions to make them more or less specific. | + | #Decide how you will respond to missing data. Not all respondents will answer all questions, which may or may not be a problem for you. <ref name="rf5">http://www.ats.ucla.edu/stat/seminars/survey_monkey.htm</ref> Ask yourself which questions are being skipped or answered incompletely, if at all. This may be due to the order of the questions, the wording of the questions, or the subject matter of the questions. If missing data is important, consider rewording skipped questions to make them more or less specific. |
#Review what sort of feedback you are receiving. See if you have unusual trends in your data and decide if this reflects reality or if this is due to a flaw in your questionnaire. For instance, your closed-question answers will limit the sort of information your respondents can give you. Your answers may be so limited as to make strong opinions look the same as weak opinions, or may not provide for a full range of reasonable answers. | #Review what sort of feedback you are receiving. See if you have unusual trends in your data and decide if this reflects reality or if this is due to a flaw in your questionnaire. For instance, your closed-question answers will limit the sort of information your respondents can give you. Your answers may be so limited as to make strong opinions look the same as weak opinions, or may not provide for a full range of reasonable answers. | ||
#* For instance, if you are asking respondents to rate an experience, you should provide them with the option to respond with "very dissatisfied" as well as "very satisfied," and many options in between. | #* For instance, if you are asking respondents to rate an experience, you should provide them with the option to respond with "very dissatisfied" as well as "very satisfied," and many options in between. | ||
| Line 52: | Line 52: | ||
*You may choose to add an "I don't know" response to questions that respondents may not have an honest opinion about. This may help you avoid collecting inaccurate answers. | *You may choose to add an "I don't know" response to questions that respondents may not have an honest opinion about. This may help you avoid collecting inaccurate answers. | ||
*Be strategic about selecting your respondents. No matter how well your questionnaire is designed, your results will be less useful if your sample is biased in some way. For example, giving a survey over the Internet about respondents' computer usage could result in drastically different data than when you give the same survey over the phone, since your sample may be more familiar with computers. | *Be strategic about selecting your respondents. No matter how well your questionnaire is designed, your results will be less useful if your sample is biased in some way. For example, giving a survey over the Internet about respondents' computer usage could result in drastically different data than when you give the same survey over the phone, since your sample may be more familiar with computers. | ||
| − | *If possible, offer something in return for the completion of the questionnaire, or let the respondents know how their answers will be used. Such incentives can be motivating for respondents. <ref | + | *If possible, offer something in return for the completion of the questionnaire, or let the respondents know how their answers will be used. Such incentives can be motivating for respondents. <ref name="rf2" /> |
== Related Articles == | == Related Articles == | ||
