|
|
 |
|
| Planning a Survey? |
For more information
contact Lee Herrin
(250) 356-0025 |
| |
 |
|
|
|
If your branch is planning a survey, BC Stats has developed a list of 20
questions which you should try to answer before you begin:
Before you start you need to ask:
1. What are the objectives of the survey?
2. Are there other sources of data I could consult before carrying out a survey?
3. How will I ensure that those who have a stake in the outcome of the survey support it?
4. How will I develop the list of people/organizations to be surveyed and how reliable is the contact information?
5. How should I design my sample to minimize cost and maximize the accuracy and flexibility of the results?
As you develop the questionnaire, you need to know:
6. What information is required to meet the needs of my project?
7. What is the best way to word questions so that I will get unbiased responses?
8. How will I design the survey to ensure the questions are clearly understood and answered properly?
9. What is the most reliable and cost-effective method of delivering the survey?
10. How should I pre-test the survey instrument? When should I use the results of the pretest?
When dealing with the respondents, you need to answer:
11. How will the confidentiality of the responses be protected?
12. Who will respondents contact when they have questions?
13. How long do respondents have to respond?
14. What will I do if they don't respond?
15. What is an acceptable response rate? What will I do if my response rate is unacceptably low?
Before you analyze and present the information, you need to know:
16. How will I assess whether or not the responses are biased and how will I correct my data for bias if it exists?
17. What techniques will be used to impute, estimate and weight the responses to give accurate, fully representative results?
18. What techniques will I use to analyze the data?
19. What is the best way to present the data so that my audience grasp the importance of my findings?
20. How will I demonstrate that the results are statistically valid, accurate and reliable?
|
|
|
|
|
