Editor’s Note: This is part one of a two-part guest post on survey use in libraries by Celia Emmelhainz.
Surveys are everywhere. You go to a government website, a vendor’s blog, an organization’s page, or step into a building: “We just want a few minutes of your time.” A scattering of survey requests linger in my email: ACRL, RDA, data librarians, IndieGoGo, four campus programs, the International Librarians’ Network, Thompson Reuters, and Elsevier. And that’s just the past month!
Then, when you try to actually open a survey, there are tiny little buttons: you have a large screen, but you can’t manage to hit any of them. There are pages and pages of Likert scales. Do they want your life’s story, told in rankings of five items and slider bars? They definitely want you to brainstorm for them, but who has time to think of the top 15 libraries in the world, ranked by specialization?
On Using Surveys Well
If I sound skeptical of surveys, it’s because I am: People are over-surveyed. Organizations repeatedly survey-blast the same users, not caring about the value of each person’s time. Samples aren’t representative; results aren’t analyzed—we just present pie charts and summary graphs as if that’s all we can do. We use them to justify our existence, not to understand the word or improve services. In the hands of the wrong person, surveys can be deceptive tools.
And yet, I find mixed-method surveys to be tremendously useful for librarians, particularly if we’re exploring a new area on which there’s little to no data in the existing LIS literature. As Dwight B. King, Jr. writes for librarians:
Focus groups are effective in drawing out users’ true feelings, but because the group is small, it is difficult to make generalizations… Interviews are good for obtaining in-depth information, but… can be very time-consuming. Survey questionnaires are often the best choice for ‘an economical method to reach a large number of people’ with a large number of questions.”
So, surveys: use them with care. Make sure they’re necessary, and well used. Ideally we should be moving to well-designed national surveys on library issues, at no cost to local libraries, plus occasional targeted surveys at the local level.
But there is still a role for local surveys. And so, I’ll talk here about how I’ve used various survey tools in libraries, and end with some advice for when you create your own survey.
Choosing a Survey Tool
I’ve worked with SurveyMonkey, LibSurveys, SurveyGizmo, Google Forms, and Qualtrics. Most have a free/student option or trials, but institutional accounts offer many more features.
Google Forms: Free to anyone with a google account. It’s easy to create forms in Google Drive. I’d use short Google Forms to gather librarian preferences on an issue, as a pre-survey for library instruction to gauge student interest in various topics, or for thoughts from people who are using our trial databases. You’re not going to be able to do a lot of analysis, so keep it short and sweet, and download a summary report in PDF. You can also send responses to Google Sheets to analyze, and/or download to Excel from there.
SurveyMonkey: I’ve used the free accounts, which allow 100 responses, as well as paid accounts. This is a great tool if you’re starting small, and just learning to design and analyze surveys. I’ve used a paid subscription to survey different sets of students or faculty, and have also used it for pre/post surveys of library instruction. It’s easy to filter results by date and only download the responses you need, so you can e.g. put a feedback form and just select the current day’s batch to download.
LibSurveys: As part of LibApps, Springshare offers LibSurveys, including both simple forms and longer surveys. The interface is meant to be simple, but adding and adjusting fields (questions) is a somewhat buggy process. Once you’ve collected responses, you can view answers by question and download to CSV. Play with it if you’ve got access to it, but let me be honest; it’s not my fave! I’d rather see Springshare integrate with one of the other survey options listed here.
SurveyGizmo: This is easily my favorite. I’ve surveyed students, teachers/faculty, and librarians at school and university libraries, done usability surveys for websites, collected reference data (before I had access to Springshare), and even surveyed 385 young recent MLIS grads about their experiences in the job market last year. I find the interface and layout attractive and easy to use, and the reports and exports also easy to use. For more advanced users, you can clean data, code textual results, and even analyze data online using cross-tab reports.
Qualtrics: The institutional subscription is wonderful but expensive, so you won’t be using it unless your library has access to a university subscription. This is a sophisticated piece of survey software that allows for detailed ‘skip logic’ (adjusting the next questions based on prior responses, to keep all the questions relevant) and survey layout. I’m just getting started in using this, through a Qualtrics working group on our campus.
If you’ve used surveys, I’d love to hear in the comments about which tools or projects have and haven’t worked for you!
Celia Emmelhainz is the social sciences data librarian at the Colby College, and leads a collaborative blog for data librarians at databrarians.org. She has worked on library ethnography and survey projects, and currently studies qualitative data archiving, data literacy, and global information research. Find her at @celiemme on twitter, or in the Facebook databrarians group.
Ruth Collings
I agree with you that “survey fatigue” is definitely a problem for most library patrons. As a psychology undergraduate I was taught in my ethics course to consider the thought and effort you are asking your participants to give you against the value you expect from the results. If you consider things this way, a poorly-designed survey is actually harmful because there is no good result to justify the participant’s work.
I have used most of these options. Google Forms is very quick-and-dirty and really is more useful when you are gathering information like attendance at an event rather than anything you want to run statistics on. And big, textual datasets still crash Google Sheets sometimes.
LibSurveys I honestly gave up on pretty quickly. Until it gets much more robust and reliable it’s not really worth struggling with.
Blackboard and Moodle both have survey options integrated and work fairly well for obvious situations like communicating with a class. The output, however, is often difficult to manage and there are no analysis tools built in to the system that make sense outside of a quiz/test situation.
I haven’t used SurveyGizmo so now I’m going to try it out.
SurveyMonkey has limited functionality in the free version, but the paid version is quite robust. Its logical operations actually work and the output is usable. The question for most Canadian researchers is if they are legally required to keep their research data off of American servers.
Qualtrics is the most expensive by far, and definition offers the most options. It requires some learning to use to it to full effect, but its documentation is good. The only complaint I would have (aside from $$$) is that it is clearly designed for more marketing/qualitative research than scientific studies.
Celia Emmelhainz
Ruth, thanks so much for your thoughtful comments! It’s great to hear what has and hasn’t worked for you; I didn’t even realize Blackboard and Moodle have survey options, so that’s something for me to check out. As you point out, there aren’t a lot of great options for libraries that are both able to handle large, complex surveys and yet remain low-cost.