Online Surveys Best Practices

Creating an online survey might seem simple at first. The latest internet survey software makes it easy to construct an online questionnaire and chart the responses. But this process is not really as simple as it seems. To get the most useful results, you need to become familiar with online survey best practices.

What Do You Need to Know?

The first step in creating an online survey is to decide what you want to know. What is the main thing you want to discover? Start by brainstorming everything you hope to learn from the survey. Then prioritize the topics you have listed. This process will help you decide how much attention to devote to each item in your online questionnaire.

In deciding what you want to learn from your survey, you need to think about how you are going to use the data you collect. There is not much point in gathering information if you don’t plan to do anything with it. What actions are you prepared to take to address the responses you receive? Will you create new procedures, start a new training program, and change your marketing plans? To make the most of a survey, you need to ask about subjects you can do something about.

In some rare cases, you might decide that the best response to what you learn from your online survey is to do nothing at all. For example, if you learned that all your employees were completely satisfied with their current health care benefits, you would probably decide to leave well enough alone. Unfortunately, such situations are rare. If you thought everyone was happy with the way things are, you probably wouldn’t have bothered to conduct a survey in the first place.

Brief is Better

People have a limited attention span. Ideally, an online questionnaire should take no more than five minutes to complete. In a typical online survey, people answer an average of five multiple-choice questions a minute. This means that you should probably limit your survey to no more than twenty-five questions. (In building online surveys, remember that one open-ended question equals three multiple-choice questions.)

People are more likely to complete a lengthy survey if they are expecting some kind of reward for making it all the way to the end. For example, people might be willing to answer a long list of questions about their health if they expect to be told at the end how long they are likely to live.

To determine how long it will take to complete your survey, try it out on five or six people. Don’t tell them they are being timed. If you do, they might run through the questions faster than they normally would. If some people take much longer than others, ask them if there were any questions they didn’t understand.

Make Every Question Count

To get the most out of your survey, you have to make every question count. Don’t be satisfied with your first draft. Once you have written a question, try asking the same thing in three or four different ways. This exercise forces you to decide if your first attempt is really the best way to get at the information you want to know.

Once you have written a few different versions of a question, you can either pick one or combine parts of several to create the best question possible.

Your questions should be clear, brief and easy to read. Avoid unusual vocabulary or jargon that some people might not understand. Keep sentences short and simple.

Beware of Bias

If you really want to know what people think, you need to be careful not to steer them in a particular direction with your questions. The most obvious example of biased questionnaires is the "push polls" used in some political campaigns. A push poll question might ask, "Do you think that Candidate X’s plan to raise taxes will place an unfair burden on the middle class?"

The intent here is not to solicit anyone’s opinion on the candidate’s tax plans. The intent is to suggest that the candidate’s plans are unfair.

Push polls are obviously slanted, but bias can sneak into your online questionnaire without you noticing it. Sometimes "loaded" words can have a strong influence on how people respond to a question. For example, if you ask people whether the check-in procedures at your hotel "should be improved," they are likely to say yes. But wording the question this way doesn’t really tell you how happy or unhappy people are with the current procedures. Instead, you might use a rating question, such as, "Please rate the convenience of our check-in procedures." Possible responses could range from (1) "Inconvenient" to (5) "Very Convenient."

Be Specific

Sometimes questions are so general that they are not really useful. For example, consider the question, "Do you like salmon?" What does this mean? Do you like the taste of salmon? Do you like the health properties of salmon? Do you like the price of salmon compared to the price of other fish? Do you like salmon more or less than other fish? The question is so general that the responses you get will not really tell you much.

Also be careful to avoid "double-barreled" questions, such as, "Which of these vehicles do you consider to be the safest and most economical?" Here you are really asking two questions. People might consider a particular vehicle safe but not economical, and vice versa.

Try Different Question Types

In building an online survey, don’t limit yourself to a particular type of question. The types of questions you choose will depend on the subject and the kind of information you want to collect.

Multiple-choice questions are common in online questionnaires. They can be answered quickly and they make it easy to collect and compare data. But sometimes you don’t want to limit responses to four or five choices. In those cases, you might ask an open-ended question, such as, "How can we make visitors to our facility feel more welcome?"

Sometimes multiple choice questions allow the respondent to choose multiple answers. For example, "Which of the following products are you likely to buy in the next year?" These questions can give you more useful data for marketing purposes because they are a more accurate reflection of actual consumer behavior.

You might also try using ranking questions, such as, "Rank the following five vehicles in terms of overall value." These questions can give you a useful picture of how a product or service matches up with the competition.

Rating questions typically use a tool called a Likert Scale to create a picture of how people feel about something. For example, a Likert Scale question might ask, "How important do you think it is to have a fingerprint security reader on your laptop?" Possible responses might range from (1) "Not Important at All" to (7) "Very Important."

Matrix questions combine two or more variables. For example, a matrix question might ask people to rate five different online travel services on a scale from (1) "Inconvenient" to (5) "Very Convenient."

Basically, the type of questions you use will depend on what you want to know.

Question Branching

Question branching allows you to build flexibility into your survey. Here’s how it works. A question might ask, “Do you own your home?” A “yes” response will take the person to a new set of questions about home ownership. A “no” response will take the person to a new set of questions about renting a house or an apartment.

Branching simplifies questionnaires because respondents don’t see questions that are not relevant to them.

The Online Advantage

Online surveys offer a number of advantages over surveys done over the phone, by mail, or in person. They can include a larger sample size at a reasonable price. They make it easier to collect, analyze and present data. And they allow you to use video and images in a way that is just not practical in other types of surveys.

In recent years internet survey software has become very sophisticated. It can help determine proper sample size and evaluate the statistical significance of responses. It can also make it easy to write new questions or modify existing questions from a library of online questionnaires.

The best online survey software makes it easy to create attractive, user-friendly formats. You don’t need a background in web design or graphic arts to create a polished, professional questionnaire.

What Kind of Response Can You Expect? The response you get to your survey depends mainly on the type of survey you are doing and your target audience. For example, if you are doing a customer satisfaction survey, you are likely to get more responses from people at both ends of the satisfaction spectrum – people who are very satisfied and people who are very dissatisfied.

In surveys aimed at a broad population, your response profile will probably mirror the profile of internet users in general. In broad terms, internet users are younger, somewhat more affluent, and more tech savvy than the average adult. They are about evenly divided between men and women. In terms of residence, approximately 54% of internet users are suburban, 30% are urban and 16% are rural.

If you are interested in narrowing this profile, you can include in your survey a few simple questions about age, gender, residence, etc.

Analyzing Survey Responses

Internet survey software makes it easy to analyze responses. The best packages offer a broad range of statistical analysis tools that are simple and easy to use. Advanced users can customize the analysis by modifying the default options to suit their needs. Here are some of the analysis tools that are currently available.

Frequency Distribution simply tells you how many people chose particular responses to a multiple choice question.

Survey Cross-Tab Analysis allows you to see how responses to one question affect another. For example, you could see how what percentage of people who responded “college graduate” also responded "employed at the same job for more than five years."

Average by Category lets you compare the average response for different categories of people. For example, you could compare the average annual income range for men and women, or for high school graduates and college graduates.

Cross Tab Means is a more sophisticated measurement that allows you to see the relationship among three variables. For example, you could see how highly SAFETY is rated by WOMEN who own a FORD.

Post-Stratification allows you to adjust results to reflect the true population. For example, suppose that 25% of respondents say that they own more than one computer, but you know that the average age of the respondents is 27. Using Census data, post-stratification can give more weight to the responses of older people, so that your survey will be a more accurate reflection of the true population.

Segmentation identifies groups of customers who share similar needs and who demonstrate similar buyer behavior. This allows marketers to target different campaigns to different groups of customers based on their survey responses.

Gap Analysis allows a business to see how large a gap there is between what it is currently offering and what its customers want. For example, a survey might ask customers of an auto repair chain to rate four different things: service, quality, value, and reliability. Survey results show that "service" is consistently rated lower than the other factors, so this is the area that offers the greatest opportunity for improvement.

The results you obtain with these tools can be presented in tables, charts or graphs. These displays can be customized to present data in a way that will be most meaningful for everyone who needs to use it.

A Survey for Every Purpose

Online surveys are used extensively for marketing and market research. The online approach works particularly well for marketing surveys because it allows a business to reach a large number of respondents easily and economically. Market research software also makes it easy to analyze responses and adjust marketing plans accordingly.

Online questionnaires have also proved extremely effective for customer satisfaction surveys, advertisement effectiveness surveys, and product evaluation surveys. These tools give businesses a wealth of information about who their customers are, what they want, and how they decide what to buy.

Businesses also use online employee satisfaction surveys to learn how their employees feel about everything from pay grades to parking.

Online Survey Best Practices Yield Best Results

To get the most out of online surveys, follow the best practices outlined here.

• Begin by giving some careful thought to what you need to learn from your survey.




• Keep your survey brief; make every question count.




• Beware of bias.




• Experiment with different kinds of questions.




• Utilize the powerful analysis tools available with online survey software.

In summary, always remember... online survey best practices yield best results.