Selecting online survey tools
The University of Melbourne does not currently have a standardised toolset for running web surveys, either for academic research or for business/administration purposes.
If you need to run an online survey as part of your University work, this page will give you some ideas about how to choose the right tool for your particular needs. The information on this page is based on the 2006 experiences of several departments in the University. It is not exhaustive, and your experience may differ.
Local experience
Based on both anecdotal and more principled (formal) evidence, there are a range of web survey tools in use in various locations around the University, and for scales from a handful of respondents to thousands. Generally these cluster into six categories, as shown in the table below.
Type |
Example |
HTML forms |
HTML form with either mailto:, CGI-Mailer, or CGI-Mailer + script/macro |
Proprietary, off the shelf |
Dipolar |
Open Source, off the shelf |
PHPEsp, PHP/MySQL, Perl/MySQL |
Homegrown |
ColdFusion-based with MS-SQL |
LMS-hosted |
Survey functions within WebRaft, Blackboard, available both to restricted student cohorts and more generally |
“CMS”-hosted |
Survey functions within MySource Matrix, Drupal, Plone, Zope, Joomla |
External Provider |
SurveyMonkey, Zoomerang |
Some questions to ask
In considering which type of solution to adopt for your particular needs, it may be useful to ask the types of questions outlined below. This list is not exhaustive, but forms a good starting point for making your choice.
Question types |
Comments |
Customisation |
To what extent can a corporate style (ranging from logo embedding to full CSS) be applied to surveys? |
Accessibility |
Are the surveys compliant with web accessibility standards? Do surveys work on a variety of browsers and platforms and devices? |
Publicity |
Does the service offer distribution lists for managing survey respondents? How much flexibility is there in importing lists, dynamically grouping invitees etc? |
Access control |
Can access to web surveys be controlled – by invitation code, by central authentication services, by IP restriction, - or are all surveys open to public? If surveys are authenticated, are specific respondents anonymised or traceable? |
Hackability J |
Despite survey container design, can other elements eg raw HTML or JavaScript be entered into form publishing module? |
Scalability |
How many surveys? How many questions per survey? How many users per survey? How long will surveys run? How many surveys simultaneously? What system constraints need to be taken into account? |
Compatibility |
Where will the software be installed and what platform or database constraints need to be taken into account? |
Cost |
How much does it cost? For off-the-shelf solutions there is either a licensing or time commitment cost, or both. If external providers are considered, then the pricing model also needs to be investigated in more detail: providers are differentiated by their pricing structures eg fee per survey, service fee per month, fee per respondent etc. |
Skills required |
Programming, web design (including assessment of the survey's usability and accessibility), survey design, data analysis, IT security, business processes for analysing and storing the data collected in the survey, legal obligations about information privacy... |
Open issues
Comments from University staff about these issues are welcome: please email web-info@unimelb.edu.au with your ideas and suggestions.
First, is there a general need for a standardised web survey tool (or set of tools) to be adopted and made available University-wide? Is there a common range of functional requirements that such an enterprise-supported toolkit should satisfy?
Assuming there is a real need for such a service:
- what steps could the University's web community take to help make that service a reality?
- what would be the best costing model for making the survey toolkit available to faculties, departments and administration units? Is there a difference between “business” and “research” usage of web survey tools, and do the associated costs (and cost recovery models) differ?
If there is not a need for an enterprise-supported tool, what is the best way to pool our collective knowledge into a single location, and identify best practices in using other web survey tools?
What complementary services are required to make best use of web survey tools (eg basic statistics), and how do we make these available to the University's web community?
|
top of page
![The University of Melbourne [logo]](/webcentre/template-assets-custom/images/custom_logo_crest_comb.gif){kind=logo}
