How to Make a Survey
The good news is that thanks to the Internet, online surveys are now quick and easy to design, simple to deploy, easy for the respondents to complete, the collated data is compiled in real-time and the data is easy to analyse.
Surveys can be used for a wide range of reasons, including customer and employee satisfaction, market and product research as well as marketing and promotion. It is difficult to generalise on what makes a good and effective survey, however it is fortunate that the basics are the same, no matter what the topic.
The Goals and Objective
It is important to establish from the very start what the goals and objectives are. This will help determine what questions are asked and the style of questions and answer options.
In considering the objectives it is equally important to consider how the collated data is to be analysed.
For example if an option is required to be able to analyse the collated data by age, by gender, by location it is important that this information is collected as part of the survey.
It is also better to think ahead and determine as to how the collated information is to be grouped; for example, instead of asking a respondent their specific age it is much better to ask them to indicate the age group they are in.
This will serve two purposes, it may allow those respondents who would prefer not to declare their exact age enough comfort so that they are happy to indicate their age group, but more importantly, as the data is already grouped in the age groups that are required for analysis it makes the analysis much easier.
It is very common to see examples of surveys that ask respondents to register their satisfaction using a scale of 1 to 10. There is nothing specifically wrong with this but in most cases it will be found that when it comes to the analysis of the results, there is the requirement to group the results into something more meaningful such as 'excellent', 'good', 'okay', 'poor' and 'very poor'.
It therefore follows, that if the scale is eventually going to be replaced with these more meaningful headings, then dispensing with the scale from the start and using the headings instead makes everything, including the analysis, much easier.
A World Full of Grey
Decision making is often not clear-cut, sometimes the least worse option has to be taken. Although it is often common for respondents to be asked a question and given the answer options 'very good', 'good', 'okay', 'poor' and 'very poor', if a decision is to be based on the feedback, it is quite common for the analysis to lump in the 'very goods' in with the 'goods', and the 'very poors' in with the 'poors'. Sometimes it can be argued that it is better to cut to the chase and the respondent should be asked if they are simply in support or against the proposal.
A modern approach would be to use a Slider format where the respondent is provided with two answer options, such as 'love' and 'hate', 'like' and 'dislike' or 'agree' and 'disagree'. The respondent is then asked to move a slider button to indicate how close they feel to either answer option.
Improving the Question
If a respondent is asked if they agree to a statement that 'safety is paramount' they are more than likely going to agree, why would anyone be against improving safety?
However, modify the question along the lines of, 'Our current safety record is one serious accident in every twelve thousand hours of operation, would you agree to a 20% increase in fares to help us further improve our safety record?'.
In this second example the question has been made 'real world' and the respondent is being asked to decide if they feel that the current level of safety is acceptable and if not, if they consider that improvements to safety are worth a 20% increase in their fares; the results from the second question format is likely to be very different to that of the first.
Closed questions, where a respondent selects the answer option from a list of possible answers, are much easier to analyse than open questions where the respondent is given free rein to write whatever they would like.
There are times, such as when asking a respondent for their comments on a specific subject, where an open question is essential, but where there is the option, it is better to take the opportunity to make a question closed.
Using an online survey website, such as Survey Galaxy, creating a survey is now a doddle. Add items, edit them, move them around, delete them and include images, a logo or a video.
A much unappreciated benefit of conducting surveys online is that they can be tested a lot more thoroughly than those conducted using a printed survey, and in addition the testing can be followed through to the analysis to ensure that when the results data is analysed it will achieve the survey's specific goals and objectives.
Keep in mind that survey questions should be kept to a minimum, limit the questions to 'need to know' and exclude 'nice to know' questions. Include demographic questions to help with the analysis of the results but, only ask for information that is essential to the analysis.
Respondents who participate in surveys are investing their time. Even when there is an incentive being offered to encourage respondents to complete a survey it is better to keep the survey compact, and to the point, so as to minimise the number of incomplete survey responses.
New Dawn Rising
Despite a number of excellent online survey websites having been around for a number of years, and although many people have in that time discovered the benefits they bring, it is nevertheless true to say that we have yet to brush the surface as to how online surveys can be used.
In the past the benefits of conducting surveys were rarely questioned, however and quite rightly, the time and effort required was. Now online survey websites have removed nearly all the negatives, and also added a few more positives, making them a valuable business tool.