Survey Tools

Approved: 
Yes

There are a large number of tools available for doing surveys, some of which are listed here.  Features to consider include:

Does the software run on a public website, or do you need to install it on a server within your organization?  There are pros and cons of each method.

  • If the software is running on a public website, you avoid having to set up and configure the software, something that will be particularly important if you don't have access to technical expertise.  Also, if you only run surveys infrequently, maintaining your own survey software is impractical.  The downside, however, is that in most cases, if you use a public website your survey URL web address will be that website's address, and not one showing your organization's name.  This can result in lower response rates than if the URL points to a recognizable and respected organization.
  • The converse of this is true for installing your own software: by hosting it within your organization, potential respondents will see a URL for your organization, which can increase responses if your organization is well known.  This will, in most cases, be more costly in terms of licensing software and paying staff to maintain it than using a public solution unless your organization does regular survey work.
  • Confidentiality is also a consideration, and your organization's policies about protecting data may influence your decision.  If you are using a public service, you should check their data protection policies.  Often, these sites will have quite detailed plans in place to protect your data.  Whether those assurances are sufficient, or whether you need to host data on your own organizational servers can influence your decision.

What are the main design features you require?

  • The ability to specify skip patterns is a powerful feature becoming increasingly common in survey software.  Skip patterns can often drastically reduce the time it will take respondents to complete your survey and, as a result, can often decrease your drop-out rate.  An example of a skip pattern is asking a question such as "Have you used the search feature of this resource?"  If you then have a short series of questions about search, if someone answers No, then a well-designed survey will jump right past those questions without ever presenting them to the respondent.  A poorly designed survey will either go ahead and ask them a series of irrelevant questions, or just include a text instruction that instructs the user to manually skip ahead to a particular question.  While this was common when most surveys were done on paper, it is much better to do this automatically as it will make your data much cleaner, and make the survey appear shorter to respondents.
  • Available question types is also a consideration.  Will your survey mainly be simple responses, or do you need some sort of complex question types such as multiple response tables, tables partially displayed based on skip patterns, or the option to allow for unusually long text box responses?  If you plan out your survey in advance before setting it up in the software, you will have a better idea of the functions you will need.
  • How long does your survey need to be, and how many responses do you anticipate?  Most of the tools listed below offer basic packages, which are often free but are limited in the number of questions you can include and/or the number of responses you are able to receive.  For a fee, however, you can upgrade to more fully featured packages that allow longer surveys and more responses.

What distribution method you would like to use will also influence your decision.

  • You might choose to simply send the survey URL to e-mail lists, post a link on your website, and include it in a newsletter or other communication.  This method is the fastest and least time intensive method.  The downside is that it is the least scientific, because you will have no way of knowing your response rate (ie., the percentage of people who upon seeing the invitation decide to complete the survey) or of knowing whether your sample is representative of any more general population.  If your survey is exploratory, these limitations may be fine.
  •  A more robust method is to use a known list of e-mail addresses and to use survey software that supports sending out invitations and tracking responses.  Using this method, each potential respondent is sent a unique URL, and the software will be able to track whether they have responded to the invitation or not.  This method will let you know what sort of response you are getting to your survey.  A general rule of thumb for online surveys is that 20-25% response rate is fairly normal if your list is targeted at an interested population.
  • A hybrid approach may also be appropriate, where you have a target list that you invite to participate, but also have a separate URL for any interested parties who did not receive an invitation.  If you use this method, you may want to include the more general URL at the bottom of the invitation e-mails with a message such as: "The link in this message can only be used once by the person invited to participate in this survey.  If you know of others who may be interested in participating, please ask them to use the following general link to the survey to complete their responses: http://survey.com/survey

Available tools

There are any number of available tools for doing surveys.  The following list is not meant to be comprehensive, but represents some common tools that are available.

  Public website or server-based? Invitation and Response Tracking  Comments
Lime Survey1 Server  Yes Lime Survey is a popular open source survey tool.  The software is free, and relatively easy to use and full featured.  The downside is the need to host the software on your own server, which may be overkill for simpler survey needs.
SurveyMonkey2 Website  Yes SurveyMonkey is one of the most common public survey tools.  It offers many features for free or a low cost.  There is one potential downside to its ubiquity and ease of use: some users may have already been sent a number of poorly designed surveys from this site, and may judge yours on that basis.
Survey Anyplace3 Website Yes Survey Anyplace focuses on mobile and offline surveys, which is useful for reaching mobile audiences. Note that students can request a free subscription for the Professional Plan by emailing info@surveyanyplace.com.
PsychData4 Website No PsychData is particularly appropriate for sensitive surveys, as they are designed to help ease the institutional ethics board review process.  The software also optionally allows users to register using their e-mail address, which will let them save their responses and return at a later time.  Skip patterns include the ability to link multiple surveys together based on previous answers.
 Snap Survey5  Server  Yes Snap Survey is a fully featured suite of tools with a range of options.  It is well suited to an organization that intends to do extensive and repeated survey work.
QuestionPro6 Website Yes QuestionPro is relatively easy to use, and is quite good in terms of allowing basic uses but also having advanced features available as your survey needs grow more complex.  One downside is that the URLs provided for a survey can be really long and complex, so you should probably think about having a simpler URL for respondents to follow that then links to the actual survey.