Page 1 of 3
The Trouble With Web Surveys, Written by Brian Cooper ?– Managing Director, Omnisis Ltd?
This paper examines the development of the web survey and the current problems the modern online research agency faces.
The internet has been with us for many years now, and the World Wide Web, in its present form, about 10 years. Web surveys have been with us just as long and have developed significantly during those years. The question is “have things changed for the better”? We are going to look at some of the problems with web surveys, and most importantly, what we as web survey providers feel are the solutions to those problems.
Web surveys are a key part of modern research methodology and with recent developments in technology, added sophistication in software, higher bandwidth, computer processing power, specialised internet programming languages and the skills of the web survey designers, we are seeing more and more complex surveys reaching the browsers of the humble respondent. These modern web surveys use dynamic web languages, animation, images and sounds as a way to get respondents to provide us with the answers we so desperately need. We use radio buttons and check boxes, grids and tables, drop down boxes and answer piping – all formats that you would never see on a paper questionnaire.
We are in danger of designing web surveys to suit our own egos and not for the suitability of the target audience. While there is no doubt that the resulting web survey looks good, does all the technology surrounding the visible screen actually make it harder for the respondent to firstly, access the survey and secondly, actually complete it? Let me put that another way – do you have java-script enabled on your browser? Do you use Internet Explorer? Simply by answering “no” to these two questions could result in a survey not even starting. Of course, the increase in programming and the use of images and videos increase the download time of each page of the survey - have we incorrectly made the presumption that everyone now has broadband access when statistics tell us that the level of ADSL uptake in the UK still runs at just over 50%? (National Statistics, 52.4%? June 2005). Using the latest and greatest programming techniques and software may produce the best looking and most interesting web survey there has ever been, but when it takes the recipient five minutes to access an image on his 56k modem, who benefits?
A key thing to remember about the use of surveys to collect information is that they have to be a scientifically sound way of “generalising” results to a larger population, be that a country as a whole or the users of a particular product. We need to select the members of a population carefully and give them a chance to participate, and by complicating the online survey process, this population may become too skewed to be useful.
There is no doubt that some of these advanced techniques have helped the data collection and analysis process immensely. Cleaner data means infinitely more usable results. Additionally, your respondent is more likely to have trust in the survey if logic is followed, and he is only asked questions that are relevant to him, and take into account information he has already provided earlier in the survey.