Collecting responses in an online survey
For those of us who have licked hundreds of stamps to place on envelopes that send our little surveys out into the world, wondering whether our investment is in vain, online surveys seem almost too easy. After all, aren’t a sore tongue and paper cuts from survey folding rites of passage in social research? Well, maybe not, any longer . . .
It is unlikely that social science/educational researchers and evaluators will ever get away from at least a few hard copy surveys, but use of online surveys certainly does make things simpler. As we discussed earlier, one has to be certain that the target audience is right for a technologically mediated survey, but once that is confirmed, getting it out there is a breeze. Before you pat it on the head and send it out into the cold, cruel world, though, there are a couple of other decisions to make.
1. Do you want respondents to be able to interrupt and return to the survey at a later time? Not all survey applications allow this, but some do. It is a courtesy to the respondent to allow interrupting and returning, particularly if the survey is long.
2. Is it important that only specific individuals respond to the survey? If so, you must choose a dissemination option that is restricted to that pool; you will also likely want to restrict the survey so that it can only be completed once on each computer terminal. If, on the other hand, quantity rather than specificity of response is important, you will want a dissemination method that sows widely, and a survey option that allows multiple inputs on the same computer terminal.
3. Is it important to you for the responses to either be or not be anonymous? Or perhaps, is it okay for some responses to be anonymous while others need to be known? All these options can be handled in an online survey, as long as you, as the designer, know upfront what you want.
Disseminating online surveys
So, all that being said, most online surveys give you multiple ways to launch your survey. SurveyMonkey, for example, provides the following, as described in their online documentation:
* Web Link Icon/Web Link: Collect anonymous surveys by posting a link on a website, or email it using your own email.
* The Email Invitation: Track respondents through “unique” links delivered by our mail server.
* The Popup Window: Have a survey or invitation open when people visit your website and collect anonymous responses.
SurveyMonkey documentation goes on to give examples of how each of these might meet the researcher’s or evaluator’s needs. Your decision about which of these, or combination of these, options to choose will be based on questions like the following:
1. Is your target audience already defined (specific individuals chosen because they are part of a project, have participated in something you want to evaluate, or other defining characteristic?
2. Do you have email contact information for all the target individuals you wish to survey, or if not, are you in regular snail mail communication with them?
If the answer to these questions is yes, then you can send the prospective respondents a link to the survey, either through your own email or through the survey application, or send them the link in hard copy through snail mail for them to type in to their browsers.
On the other hand, you may have a website, wiki, or blog that the target audience regularly accesses. If you do not need to limit the respondent group as tightly based on specific characteristics (perhaps because you will be collecting information on those characteristics in the survey anyway and can sort in or out on that basis), you can post a link to the survey on your website. Another option for website posting is the pop-up link as described in the excerpt from SurveyMonkey documentation above; with this option, the survey can pop up when the potential respondent accesses your site or specific areas on your site. This is a great approach for exploring how people use or value your website content.
As you can see, there are many flavors in this ice cream shop. Because these surveys are so flexible, you really want to think about dissemination; you are not as limited as you might be in a hard copy survey disseminated through the usual means.
As a final note in this post, it is important to recognize that most survey applications do not allow the respondent to print out a copy of his or her responses. In our own surveys for the NYS SPDG (the S3TAIR Project), we go into the survey, copy individual responses, and email them back to respondents for their files.
Next up: analyzing responses to an online survey