Designing surveys can be a difficult task, its not always as easy as it looks. There will always be a time when you are required to complete one, whether short or long, these will always produce excellent insights to the researcher about yourself and your opinions.
The national census is something you will have all come across. The abundance of data collected from this compulsory survey allows the government to gather a variety of information on society, primarily demographics. Of course the scale of this survey is beyond anything you may come across on a daily basis. Generally you will encounter some sort of feedback survey asking you to rate a service or company, most of which are common amongst online shopping sites.
Depending on the purpose of the survey, the design of it will vary largely, ranging from the number of questions, the length of the survey, the language used and many other considerable factors. The purpose of this article is to provide a generalised checklist of criteria which can be applied to all surveys created.
- How will the survey be conducted – This will depend on a variety of factors such as your audience, cost, time, the size of the sample and many more. Think carefully about the practicality of your chosen method, e.g. over the telephone, on the internet, through an app etc.
- Consider who your audience is – Understanding who will complete your questionnaire is one of the most vital things to consider. Of course this would be the first thing you can consider, but it is important that you understand their perspective, and that the questions you use are relatable to them.
- Language and terminology – While designing your own survey, it is easy to accidently assume your audience will understand the words used in the question as well as you do. Ensure that you use simple language that is easy to read, or at least provide definitions when including more complex terms.
- Provide clear instructions and examples if required – Once again it can be easy to assume a survey is easy to complete while you are the one designing it. If you are providing the option to select one option out 5, make sure you state that the respondent should circle the answer, or whatever your preference may be. Small errors like this can have an impact when it comes to extracting the relevant data.
- Decide on a structure – Structured, semi structured and unstructured surveys all have their pros and cons, but choosing the right one will determine the type of data you collect. The decision to receive more qualitative or quantitative data will help you to decide what structure is best.
- Make use of validated scales – You may find that other professionals have researched similar topics, and used scales to answer specific questions e.g. measuring customer satisfaction. It is worth using these scales in your survey as they have been proven to work by industry experts.
- Layout – As much as we would all love to gain as much information as possible from survey respondents, they don’t all have the time to complete a series of questions. Consider keeping your survey short as there is a good chance after 10 questions the respondent may lose interest. In addition make the presentation of the questions look simple rather than crowded, or your audience will see it as a page of clutter and be reluctant to respond.
- Point of contact – The smallest amount of detail can make the difference between someone filling in your survey rather than dismissing it as there is a lack of understanding. By providing a contact point, respondents can contact someone in regards to any queries they may have about the survey.
- Do not provide leading questions – A leading question is one where respondents are encouraged/likely to produce an answer that is wanted by the researcher. By doing this, there is an element of bias within your data and so the validity of your data is compromised. Have other people read your question first to see if there is any indication of a preference for one answer.
- Question your data – Conducting an analysis of the data collected will provide you with indications regarding the areas you are focusing on. For example if customer feedback is mainly positive in a particular store you know that there is no problem to address. To take your analysis one step further, find out exactly WHY customer feedback is so positive in a certain store. Is it the way the staff are trained, and could it be replicated elsewhere? There is always an option to conduct further, more specific research, if time and costs are not a restriction.
Bear in mind that the criteria checklist above are only a proportion of factors that should be considered when designing surveys. There are other elements that should be taken into consideration which are of equal importance to the ones listed. Ensure that you conduct a pilot study where a small sample of your audience is able to complete the survey and provide you feedback for potential areas that require redesigning.