University of Missouri Extension - Ste. Genevieve County

1. PURPOSE -- What do we want to know, and what are we going to do with the answers? People often fall into the trap of doing a survey because it’s a familiar and common way of getting information, but it may not be the right tool for getting the particular type of information you’re seeking. Start with figuring out the purpose of what you’re trying to learn, and clearly identify who you want to learn it from - your target audience.

2. DESIGN -- What procedures and methods are the most effective and the most feasible for our needs? Now that you know you want to do a survey, explore which type will best reach your target audience while meeting your needs. Options include a telephone survey, having a table in a public place (or a place where your target group will gather), mailing the survey out, putting it in a newsletter, face-to-face interviews, and more. This is also the time to decide how many surveys you can handle. Depending on the size of your target audience, you may need to use a sample of that group. If you mail out your survey or use another mass production method, keep in mind that you’ll only receive 10-50% of the surveys back – but don’t let that deter you if it suits your needs.

3. PLAN -- What do we need to do our survey? What is our timeline? How are we going to finance the costs associated with the survey? Think through each step of implementing your survey - getting the survey out, getting it back, tabulating and analyzing the information, and preparing the final report. What do you need to accomplish each step? For example, if you’re using volunteers, what orientation or training will they need? If you’re doing a mass mailing, how much will the postage cost? People are often the most unprepared for dealing with the completed surveys, and as a result they may wind up without useful information.

4. DRAFT -- What questions do we want to ask? What order should they be in? How long should it be? How can we make the survey attractive to our target audience? Now that you know what type of information you’re seeking and how you’ll go about gathering it, do some brainstorming on the questions. Don’t get caught up in trying to find the exact language or phrasing early on – get all the ideas first, then you can begin editing. Watch out for “double-barreled” questions – asking two questions in one. Also, you’ll probably want to include some demographic questions about your audience (such as age, gender, occupation) BUT only ask those that are relevant to your needs. Asking for some personal information, such as income, can serve to discourage people from completing the survey – if you don’t really need to know, don’t ask. Once you have the questions set, work to make your survey ‘user-friendly,’ meaning easy to use and attractive, so that people will want to complete it…or at least won’t mind.

5. PRE-TEST -- Did our questions obtain the information we wanted? What changes need to be made? This step is really crucial to the success of your survey, but unfortunately people often think it isn’t necessary. We start to get crunched with time and resources, and decide that we’ve put so much work into the survey instrument that we’re sure it’s fine. Imagine, however, getting your surveys back, and after all that work discovering that you really didn’t receive any useful information. Always pre-test your survey on at least ten people, preferably people from your target audience.

6. IMPLEMENT -- Are things going as we anticipated? What refinements need to be made? You’ve made it to the big moment, getting the survey underway. Watch for problems, but don’t make any major changes if you can avoid it. If you do, separate the new surveys from the first batch – otherwise your data will be “contaminated,” meaning that the change you made could affect people’s answers.

7. PROCESS -- Is the data we’ve collected accurate? Are there any surveys or questions we need to discard? Go through the surveys before you give them to whoever is going to be entering the information into the computer. Watch for missing pages, illegible writing, or other significant problems. You’ll probably notice that some people didn’t fill answer all the questions. Decide which are critical questions that must be answered by everyone (usually the demographic questions), and discard those surveys that don’t answer them. Sometimes, in spite of pre-testing, we may discover that a particular question has proved problematic – maybe it wasn’t as clear as we thought, or people didn’t interpret it the way we intended. If so, just eliminate it from your data entry. No information is better than wrong information.

8. ANALYZE -- What does the data show? What themes emerge? Did we achieve the purpose of the survey, and learn what we wanted to? For most groups during a survey, detailed statistics are not necessary. You’ll want at least “frequencies,” meaning the number and percentage of people who answered a question a particular way. Make sure that the percentages reflect the number of people who answered the question, NOT the number of people who completed a survey (remember that not all people will answer all questions). If you’re using a statistics software program, you may also want to run a few cross-tabulations – comparing the answers to two questions. For example, you may be interested in learning whether there are differences in how men or women answered a question (assuming you included gender as one of your demographic questions, of course).

9. REPORT -- Are we sharing the new knowledge with the appropriate groups or people? Is our method of reporting effective? Sometimes a survey will reveal information that may be useful to someone we didn’t consider in the planning process, or perhaps you have particularly dramatic results that you feel should be shared with a wider audience than you originally intended. It can be helpful to stop a minute, go back to your original purpose statement, and think through how you want to go about sharing what you’ve learned, and who you want to share it with. Don’t forget about the people who completed the surveys – they’re probably interested in the results as well. If your survey and report are lengthy, consider preparing a summary…which is more likely to be read than a larger document.

10. NEXT STEPS -- How do we want to use this information? What other questions does this survey raise? Now that the survey is behind you, take a little time to reflect and celebrate. While the survey process is fresh in your mind, jot down some notes on what worked well and what could have been done better, in case you want to do a survey again. Review and discuss the results, identifying new questions and priority issues. Where do you go from here? What actions will you take as a result of what you learned? The biggest complaint people have about surveys is that after they bothered to take the time to fill them out, nothing happened as a result. That can lead to credibility problems, as well as less cooperation the next time you want to gather people’s opinions. A well-planned survey provides useful information, and leads to real actions. Good luck!


Return to the Ste. Genevieve County main page
		Ste. Genevieve County University of Missouri Extension Center stegenevieveco@missouri.edu Last revised: 06/10/08