Basic structure of an effective survey
I am assuming you are thinking about doing the survey online or electronically. However, if you are thinking about the paper form of a survey, most rules apply. There are a lot of options available, when it comes to doing e-surveys. Just about every one of them will offer their version of 10 tips for an effective survey. Some of them are listed below (I do not maintain these links and they may break in the future):
- Survey best practive from SurveyMonkey
- Building effective surveys from Qualtrics
- Writing effective survey Qs from SurveyAnyplace
- How to write good survey Qs from HubSpot
It seems if someone or anything in this market couldn't come up with 10 tips, their product is of less value. You don't need to read through those guides. They are more or less the same thing; some of them are using the same wording and some of them do paraphrase a bit. I will summarize the relavent part in the following box.
- Ask close-ended Qs
- Keep Qs simple (don't mix multiple Qs into one)
- Give balanced choices (cover all the options uniformly)
- Provide incentive
- Test drive before deployment
If you only have 30 sec to spare, just read the words in the box above in 5 sec and think about it for the rest of the time. When you think about the survey that you did before, you just realize that most of them do not follow those guidelines. For example, most marketing surveys ask leading Qs, provide skewed choices.
- Very Good
They could put in their ad that according to an survey most customers rate their product as satisfatory at least. I think this Q is very effective for their product marketing. So, it absolutely depends on your objective whether you should follow those guidelines or go the opposite direction. Most of the time, no one will dig into those customer surveys to uncover these tricks, and those marketers get away with it.
Here, I am doing a survey for research (non-commercial). I have no interest as to what the respondants would choose as long as they pick something. So, the objective is to maximise the responding rate and to get unbiased results. It is very important to keep the percentage of close-end Qs high, but also provide chances to provide additional input (tip 1).
In the following, I will give the structure of the survey I designed.
Here ask for the email address (need validation), and the demographical Qs to guide the respondants to their corresponding section of the survey. You need to keep your Qs short, non-intrusive, and easy to do. If you think asking Male/Female Q is important here, do so but also provide some incentive, for example a free T-shirt with different designs for different sexes. This section is the gateway to your survey, don't lose your people here. Lure them in! Give them the incensive right here, right now.
Organize your Qs into different sections according to the demographics of your people and the subject matter. It would be easy for your people to answer Qs about one topic at a time. Your people don't need to go through all the sections of the survey. Depending on the answers they give to questions in one section, they may skip some sections altogether. Be inteligent and responsive here.
Say the things you need to say according to your institutional rules, like ethics, privacy and such. Then, say the incensive again. Don't let them wonder where is the T-shirt and feel cheated. Let's be honest. About 99% of your respondants do your survey just to get that T-shirt or gift card, and that 1% just don't care. Free T-shirts work perfectly for college students but become ineffective for other groups. When you have a mixed group, use gift cards instead.
The next step is to create the form and the associated files.