SURVEY MYTHS DEBUNKED - You Don't Need To Know Your Respondents
Obviously, with all the advantages online surveys offer, you shouldn’t be held back by not knowing all your respondents personally. They are out there; you are here in front of your computer. You will get to know anything you want from the answers to your survey, right?
Right. Yet you do actually need to know your respondents quite well before you begin to survey them. It can be the very pillar of your success.
Know Your Respondents
The confusion comes in because, practically speaking, it’s impossible to know all your respondents on a personal level. But that’s not necessary. What is necessary, however, is to know some particular characteristics about them. Depending on what your survey is aiming to assess, you might need to know beforehand where they live, what their income level is, what age they are, their level of education, their marital status, or their religious/political/cultural beliefs. Or it could be something completely different like which brand of potatoes they prefer, on which day they do their shopping, or what their favorite restaurant is.
Why? Because you will probably have to do some survey sampling, and to do this properly you will need that kind of information. Also, it will influence how you can best design your survey.
What sampling has to do with knowing your respondents: as described in our article about survey sampling, there are four main types of sampling. Two are random by nature, only requiring you to choose from a list of possible respondents in no particular order or by picking every 4th person on the list. You might think that in this case you need not know thy respondents, but even here that is very helpful. Otherwise, how will you compile this list from which you're choosing?
But if you are going to use any of the two other sampling types – stratified or cluster sampling – you definitely need to know quite a bit about the people on that list. That is because you are going to divide that list into categories according to some shared characteristics, such as their age, their annual income, or the level of education they have received.
What survey design has to do with knowing your respondents: understanding your respondents is also necessary so as to enable you to design a good questionnaire. This is important because you will need your respondents to understand the questions easily, since this will increase your response rate. So in that case, their level of education or their religious/political/cultural beliefs may well be a big factor influencing which questions you are going to ask or – maybe more importantly – how you’ll phrase it.
The thing with survey questions is that there is always more than one way of asking, and there is always a risk of a question being ambiguous or having more than one correct answer. Even the order of your questions matters, because an early question can affect the answer to subsequent questions. How you phrase your questions is therefore as important as what you ask. The general rules of the game are:
Be specific. Do not just ask about sanitation products or Internet use. Ask specifically about sanitation pads or gambling sites, if that is what you’re after. That also means not using trade jargon or vague terms like usually or often or on average. Usually, that will create confusion.
Do not ask double-barreled questions. That’s questions that have more than one question in one sentence, like Do you like Internet sites that have products reviews and free trial offers?
Give respondents clears options with ranges. Like, if you’re asking about their weight or age or income, set the ranges.
How To Get To Know Your Respondents
There are three techniques of getting to know your survey respondents before starting your actual survey. These are:
Expert review: have the survey questions reviewed by an expert in the field who has experience with the population you are targeting. Depending on the survey subject, consulting an expert will ensure that your questions are properly phrased and easily understood by the respondents. Of course, you can also always ask the company you are using to conduct your surveys with.
Pre-test your survey questions: test your questions on a small group of people, such as colleagues or a select focus group. This group doesn’t necessarily need to be part of the targeted population, but they should at least know what the survey is aiming to achieve. A second opinion may be all you need to craft a winning questionnaire.
Conduct a small pilot survey: select a few of the target respondents and have them take the survey. This tactic will quickly show you how well the respondents understand your questions, the level of details coming through, and areas of confusion or ambiguity, if any.
Survey Creation Service
If you don’t want to spend too much time on all this and just get it over with, you can always pass the burden to our Survey Creation team. Our team of experts has plenty of experience in all aspects of survey research, and will happily assist you putting together your perfect questionnaire. If you want, they will also conduct the survey and analyze the results as they come in.
It's Super Simple.
Is there something about surveys that you think is confusing or often done wrong? Give us a shout in the comments and we will do our best to help clear it up!
Photo credit: Meshari Karkari (thanks, Meshari)
Jan 31, 2014