Factors Leading to the Success of Your
Survey
By Rob Donahue
[Editor's Note: Some of you may recognize Rob's name. From 2000 to 2002, he worked with Hostedware. He's currently pursuing an MBA from the Leeds School of Business at the University of Colorado-Boulder. We couldn't be happier to publish his article in Hostedwire!]
When you run a survey,
you want to get the most accurate data possible. You also want to ensure your
respondents enjoy the process so that they'll participate in future surveys.
And you want to do it all as easily as possible.
Sound like a tall
order? Not at all, if you keep these ten keys in mind as you create your next
survey. Of course, the first three keys have to do with the questions you use
(the backbone of any survey). Next, you'll learn about supporting and
respecting your audience and how that'll improve your survey experience. And,
finally, you'll learn the keys to unleashing your survey on the world, from
whom to invite and when to invite them, to how to set up your introduction
page.
Question & Data format / Analysis
Intentions
What kind of
information would you like to have when your survey is finished?
For instance, people
often use open-ended questions in their surveys because they allow respondents
to answer with the exact right response. So, for the question, "What is
your favorite color" respondents don't have to decide between blue and
green, but can enter "aqua" or "blue-green."
With open-ended
questions, you will get more specific answers. But, you'll also have a lot more
work to do in assimilating the data. First, you'll need to decide whether or
not to group similar responses. (For instance, are aqua and blue-green the same
color?)
Also be prepared to
deal with misspellings, possible profanity, and the occasional indecipherable
response. If you use open-ended questions, the analysis phase will certainly
take longer.
If the idea of
filtering through open-ended answers is unappealing, use an alternative
question format, such as multiple choice.
By determining the information you'd like to have at the survey's completion, you
can structure your survey to gather that information in the method that will be
most efficient.
Answer Scaling
When asking questions
that require survey participants to rate items on a defined scale, always
provide respondents with a sufficient range of possible answers. If you ask
respondents to indicate their level of satisfaction with a series of
statements, always provide a balanced Likert scale.
A typical answer
scaling for these question types would be Strongly Agree, Slightly Agree,
Neither Agree nor Disagree, Slightly Disagree, Strongly Disagree. Another
method of asking these types of questions is to ask respondents to rate items
in a scale of 1 to 5, 1 being "Strongly Disagree" and 5 being "Strongly Agree."
Whatever answer
scaling you decide to use, keep the scaling consistent throughout your survey.
Optional Questions vs. Forcing
Responses
Do you want to force
respondents to answer any or all of your questions? Before you decide, consider
the survey from your participant's point of view. They're doing you a favor by
providing their feedback (even if you're offering an incentive, they're still
using their own time to complete the survey).
Forcing responses to
all questions may annoy the respondent. If you force a response to questions
that are not easy to answer, provide an "elect not to respond" or "don't know"
answer selection.
You may want to
consider forcing a response only on the most important questions and leaving
others optional. This way, respondents will feel some measure of control over
their experience.
Multilingual Surveys
Always make sure your
survey is clear in each language it's being offered in. This usually means
working with native speakers to ensure there's no room for misinterpretation.
Aside from the quality
of the translation itself (which must be impeccable), you'll also need to make
sure the survey remains consistent in structure throughout so that data can be
assimilated across survey translations.
Population Characteristics
If the data you need
requires that your participants fit a specific demographic profile (for
example, if you'd like to measure political opinions of females in New York,
aged 18-45), you'll get the best results if you can filter your distribution
list to send out invitations specifically to people who fit your profile.
If you are unable to
sufficiently filter your distribution list, then your survey should begin with
"qualifying questions" to weed out respondents who don't fit your
profile.
Survey Fluidity & Flow
You'll get the best
results from a survey that leads respondents through the questions. This flow
is achieved primarily by proper survey organization. Ask your related questions
together, move logically from one area to the next, and build up to bigger
questions with smaller ones that bring participants to the right frame of mind.
A clear survey
structure, complete with guideposts (like section headers and page titles) will
help your respondents move from one section to the next. Not only will better
organization lead to higher response rates, but it will also reduce
participants' frustration with more difficult questions.
Sensitivity
Always remember your
survey may be the only contact someone has with your organization. Think of it
as a first meeting and ensure the writing throughout the survey is "on its
best behavior."
Not only can offending
a respondent damage your public reputation, but it can also skew your survey results.
Engage several colleagues in the proofreading process to ensure you haven't
overlooked any language that may make some participants uncomfortable.
Proofreading & Trial Runs
Once you've created
the survey itself, you're not quite ready to unleash it on your respondents.
Instead, take some
time to proofread the survey, and ask several colleagues to look it over.
Clarify any questions that may confuse some participants. Make sure there are
no typos that may make it impossible to choose an option.
Finally, go through a
trial run prior to inviting your target audience to participate. Ask several
people to actually go through the survey in its final format to ensure the
questions are clear and the method of delivery is flawless.
Timing of Invitations
Once your survey is
ready to go, you need to invite participants to respond.
If you're delivering
the survey invitation via email, you have maximum control over scheduling when
your participants will receive the invite. And that means you need to carefully
consider the pros and cons of specific delivery times.
Are you emailing
people at their work email addresses? And do they typically work Monday through
Friday? Then, chances are, sending the email over the weekend means more people
will overlook it (in their rush to empty their inboxes of spam).
Send your email when
people are ready to respond to maximize responses to
your survey. (Otherwise, many people will put off responding until "later.")
Think about the time
of the day and day of the week that you would like recipients to receive
invitations, and consider time zone differences if appropriate. Unless you are
up against a hard deadline, you may find delivering mid-week makes the most
sense (Tuesday through Thursday).
Objectives Clearly Defined--and
Outlined
What is a participant's
first stop with your survey? The introduction page, of course. And the right
introduction page can effectively guide participants through the survey,
clarify expectations, and make respondents feel their participation is
important and appreciated.
Make sure your
introduction page clarifies:
-
Who's conducting the survey
-
The purpose of the survey
-
How long the survey will take
-
What you expect from the respondent
-
How the data will be used
Your introduction page
should also include contact information in case the respondent runs into
technical trouble or has questions about the survey.
Another crucial
ingredient on the introduction page is your survey's instructions. Make sure
these instructions are easy to follow, succinct and cannot be misinterpreted.
Along with explaining
the actual survey-taking process to the respondent, make sure your introduction
page also clearly explains whether or not responses are confidential (and/or
anonymous).
By considering these
10 survey factors, you can create a better survey experience for your
respondents. Creating a better experience means people will be more willing to
participate in future surveys. And, it also often leads to more accurate
results.
If you're ready to create an online survey,
but are daunted by the technology involved, give us a call. We'll assign you to
a friendly survey "ambassador" who'll walk you through the technical
process a little at a time.
Discover how people just like you are
setting up surveys in just a few hours using our technology tools!
To get started, email Hostedware's customerservice@hostedware.com or call toll-free at (800) 211-6967, or (949) 585-1500.
To Top
|