communication department

Volume XX Number 1

Polling Your Members: Keep It Short and Simple

By Brenda P. Huettner, Fellow, Southern Arizona Chapter

For each issue of Tieline, a representative of the twenty STC Special Interest Groups (SIGs) shares wisdom, hints, and lessons learned on different aspects of facilitating virtual STC communities, whether subject matter- or geographically-based. Currently, SIG leaders are rapidly exploring various collaborative tools and management styles to provide benefits for their SIG members—and to recruit SIG volunteers and leaders.

As Edwin Schlossberg says, “True interactivity is not about clicking on icons or downloading files, it's about encouraging communication.” SIG leaders are working on just that. We hope that some of the information in this series will be applicable and helpful to other STC community leaders as well.

Judith M. Herr, STC SIG Advocate

All communities, virtual and geographic, exist for the benefit of their members. As a virtual community, however, it is sometimes difficult to know what most of the members will see as beneficial. When you don’t have the opportunity to regularly meet face to face with all or most of your members, you need to find other ways to gather information. One way to collect information is to conduct a survey.

Web-based survey tools allow you to quickly and inexpensively poll your members for whatever type of information you need. Currently available online survey tools make it quick and easy to turn out the type of survey you need, whenever you need it.

A well-designed, effective survey takes significant planning, design, and execution. Two very good resources are the STC Usability SIG resource page on surveys and Questionnaires in Usability Engineering: A List of Frequently Asked Questions (3rd ed.), compiled by Jurek Kirakowski of the Human Factors Research Group at the University of Cork in Cork, Ireland.

General Guidelines

Keep it short and simple! You’ll get better responses from shorter surveys than longer ones. People can get tired if a survey is too long or complex, and they might not bother to complete it. Worse, they might begin to randomly answer items just to get the survey over with, which will skew your results. As you build your questions, keep the following points in mind:

• Why are you conducting the survey? Before creating a survey, write down its goal. What is it you are trying to measure, and how will you use the information you gather from the respondents? As you formulate your questions, keep your goal in mind.
• Who will get the survey? Will you ask everyone in your community, or just a sample? Do you want responses from nonmembers? Do you want to allow people to respond more than once?
• How will the questions be phrased? For the most accurate results, make sure to keep the questions neutral. If your questions are biased, the answers will also be biased. Verify that all your questions directly address the goal of the particular survey. Unrelated questions will only distract respondents and clutter later analysis. Also, make sure to choose the appropriate question type. Most survey software packages allow you to create a variety of question types, such as true/false, fill in the blank, multiple choice, or “matrix” or “scale” questions.
• When will the survey be open, and when will it close? Most people who complete your survey will do so within a few days of receiving the announcement. Leaving the survey open for a longer time will not generate more responses unless you send out reminder messages.
• What will you do with the results? Make sure to include this information on the survey itself so that your members know what to expect. Will the results affect a specific decision (such as the topic of a workshop or timing of a meeting)? Will they be posted somewhere for your membership to see?

Before sending the survey to your whole team, run it past a small sample of respondents. A test run will help identify problem areas among the questions, validate possible answers, and check the general clarity and usability of the survey.

Survey Components

To get started, you need to decide what you want on each page. There’s typically some welcoming text at the start of a survey, and there may be specific instructions, a message from the community leaders, or other introductory information.

Then, of course, come the pages with your questions. While you want each page to be clear and easy to read, you don’t want to have so many pages that the respondents get tired or bored and cancel out in the middle. If you’ve remained focused on your original goal, you may have anywhere between three and ten total pages. You’ll also need to determine whether or not each question is required. Will the user be allowed to skip any of them?

The last part of your survey should be some sort of thank-you page. This may be hosted on the survey site—at the bottom of the last question page, perhaps—or it may be a page on your community site.

Once your pages are set up, define the other parameters of your survey. While this may vary a bit depending on the survey tool you use, basic components include:

• Open date. When can people begin to use the survey?
• Close date. When will you stop accepting responses?
• Security. Will you require a password? Will it be one password for everyone in your community, or will you require a unique password for each participant? Will you allow multiple responses from the same computer?
• Limits. How many responses will you accept?

The final step is to create the announcement and tell your members about the survey. Some tools allow you to enter e-mail lists and automatically generate the messages; others provide a link that you can send to your own e-mail list. It’s good practice to supplement the e-mail announcements in as many ways as possible. For the best response, talk about your survey at meetings, during phone calls, on your Web site, and in your newsletter. Once you know what your membership wants, you’ll be better able to provide it.

Software Packages

The following companies allow you to try out their survey packages for free.