Anne Abate will be leading our next discussion on "Surveying User Needs: Do We Really Want To Know?" Anne Abate is librarian at Dinsmore & Shohl, the largest law firm in Cincinnati, Ohio, and has held this position since 1988. Prior to working at Dinsmore & Shohl, she held professional positions in a variety of libraries. Anne holds an MLS from the University of Kentucky, and HAB from Xavier University, and is currently studying for a doctorate in computer technology in education from Nova Southeastern University in Fort Lauderdale, Florida. Her discussion topic stems from her current research on her dissertation. Please join us in a discussion of user needs.
Some library administrators emphasize the need to put users at the center of library service. (a) The needs of the user should drive both the support of existing services and experiments with new services. How can we actually tell what the user truly needs? For years, librarians have been relying on their knowledge of sources or simply on intuition to suggest what they determined to be the appropriate services to meet the needs of their particular user group. After all, how can users tell you what they need if they don't know what is available. The librarian often has been viewed as a guide or mediator between the users and the potential services. Are we really serving the needs of the users if we do not open up all possibilities for them? There must be some way to gather more open responses from the users.
However, open ended questions on library user surveys can cause problems. Do we really want to know that college students are writing term papers based upon what they read in the newspapers or see on television or bump into on the Internet? In some ways, we want to know this. But, once we know it, what should we do about it? If we ask public library patrons what they want, will we end up totally replacing our valuable book collections with videos and multimedia CD-ROMs? Further, do we have the funding to provide what the users really need, or are we just setting them up for disappointment when we cannot deliver what they say they need.
In order to find possible solutions to the user needs dilemma, perhaps we can look at other disciplines. One area where the opinion of the user has become important is in user-centered computer interface design. This recent trend has been the outcome from years of computer and machine design that has neglected the user. (b) Suddenly, designers are starting to ask the users what they want and need. This has delivered some surprising results. Early microwave ovens are a good example of design that didn't consider the needs of the users. These devices were designed with more features than necessary. The full featured control panels now are being pared down to a simple start button. The same is true with many computer programs. There may be lessons for librarians in user-centered design of computer software, hardware, and even toasters. Perhaps the changes that have been driving the move to user-centered design would also be appropriate in the library setting.
One institution that has seen this connection clearly and actually is bringing together library user needs and computer interface design is the Cornell University Mann Library. They have an Interface Designer on their staff. (c) The library seeks to package information and provide library produced pathways to meet the needs of their users. They have done many other innovative things in an attempt to bring the library to the center of the picture on networked information access.
We must explore the needs of the library user now in order to be prepared for the digital library of the future. Some imaginative and amazing projects are being carried out as part of the Digital Library Initiative. (d) These projects are seeking new ways to deliver information on a broad scale. The digital library of the future will provide access to information in ways that we cannot even begin to imagine. These future libraries may be able to provide this access with no physical collections. This will put up yet another block between the librarian and the user. Likewise, we must discover new ways to determine what the patrons actually want and need in order to utilize the new technologies that are under development for information delivery in many formats.
While I have asked a number of questions throughout this essay, here are several specific questions to ponder in the first week. I look forward to your comments.
(b) For an interesting elaboration on just how silly design can be, see Norman, D.A. (1990). The Design of Everyday Things. New York: Currency Doubleday.
(c) Garrison, W.V. (1993). Integrating Networked Information into Library Services: Philosophy, Strategy, and Implementation at Mann Library. Emerging Communities: Integrating Networked Information into Library Services. Papers presented at the 1993 Clinic on Library Applications of Data Processing, April 4-6, 1993, p. 223-240. Urbana-Champaign, IL: University of Illinois.
(d) A good interim report on these six projects was published in the May 1996 issue of Computer, published by the IEEE Computer Society.
Are you studying administrator's responses to their clientele's requests for more services which are often expressed through campus library suggestion boxes or inquiries at the reference desk?
Or are you dealing with symptoms generated by the absence of a campus library program to actively teach patrons how to work with the latest tools so that they can get more out of them?
A campus with commuting students and working adult students needs to have a different structure to its library services and how it surveys changes has to differ.Does the library know how to conduct user surveys? If you ask a body of students can we raise your fees to cover the cost of library materials? You will get surprising support but lack of input can generate negative results.
A survey for numbers to support planned changes that are of positive benefit to the users, will get entirely different results than those which diminish user access. Most users are suspicious of answering honestly or even participating in surveys that can might hinder them in their work. I have participated in surveys within public, special and academic institutions and public responses are quite surprising when the survey instrument is not well designed or well worded.
Margaret Slusser mentioned several interesting points. The survey instrument must be well designed to avoid an unexpected response. Users may not answer honestly if they think their responses will change the service that they need. It is also important to decide what you are measuring before you begin the survey process.
Although we have had a slow discussion so far, I would like to add another element to the question. How are emerging technologies going to change our relationship with the users? Will they make surveys of user needs more important? Many people consider themselves to be active library users yet never enter the building. How can we meet the needs of these invisible users?
I hope some people will be returning from vacations this week to join the discussion.
Random selection of the population from which you expect end-users to come from is essential. If you only contact persons who use your product or service then you will never know what those who might use it want and need.
Even then, you need to be vigilant because those who don't use a product or service think any survey about it does not apply to them because they don't use it!
For example, in our eLib project we randomly selected from alphabetized lists of academic staff who taught 0.2 FTE or more in U.K. West Midlands universities. We interviewed them in their offices, but to just schedule the interview was difficult because we would get the response. "You don't want to talk to me because I don't use computers and Internet. You want to talk to Professor Smith who is really up on these things. He will give you the best input." Even with careful explanations that he/she was indeed important to us because we wanted a cross-section of IT users, he/she still devalued his/her input.
I might add that some of the best response information came from those people who do not use computer networks. They were in many ways more realistic about what could be achieved than the converts who gave back the assumed benefits of Internet in library services.
There is substantial evidence, I believe, that people (of all kinds) have been getting progressively fed up with all kinds of surveys since about the 1940s. This leads to low response rates, which makes the validity and reliability of the information suspect. This phenomenon is at least partly responsible for the gradual shift from door-to-door surveys, then mail, then phone, and now, possibly E-mail.
There is some evidence that people are most likely to respond to surveys when they feel their answer will make a difference in something they care about. (And some evidence that people with complaints are more likely to respond).
I think the above suggests that libraries need to be sure anyone cares, and that they need to be very, very careful in question wording, instrument design and the like. This might well suggest that the average information center ought to hire an expert, rather than doing it itself.
In the long run, I expect we will be inundated with ever more junk email, as we have been with junk mail. Thus, many potential respondents will not answer, because they delete the stuff automatically.
On a related point, I seem to recall that a number of White House Conference user surveys found that although many people agreed that various services ought to be part of their libraries, many were unaware that the services were already being provided. This would suggest a major marketing blitz, which has never happened.
I personally, tend to read over any survey I get very quickly, and avoid those from commercial operations, or in which I have little interest, or where I feel that the questions are silly/unanswerable/etc. But then, I may just be a curmudgeon.
As with most survey data, the results only tell part of the story of what users want and need. Reading through past surveys I have conducted in libraries, and the literature reporting survey results, I am struck by how much information on user needs is not there. We always asked afterwards, "why did they say that?" or "what really happened to this person that led to saying that?" etc. This is not merely a problem of writing decent questions and testing them on a sample. It is a function of the complexity of what we or they mean by user needs.
One solution to this problem has been to get much closer to users and to study what users actually do, and discover user needs in field data. Surveys are by definition retrospective. As the literature on retrospective reports shows, people do not have good memories of processes they have engaged in in the past, near or distant. Thus, surveys as a method for determining user needs are inherently weak. Concurrent reports provide grounded, situated information about user needs and reveal user needs in process.
The field data I have collected shows that user needs are complex and cover several areas of "need." It reveals affective, cognitive, and sensorimotor information needs of academic library users by either tape recording (with permission) or by assisting, observing, and manually recording such things as the questions they spontaneously ask while using a system, the comments they make about the process itself or about the electronic environment, the errors they make during the process, etc. If one is with the user while they are working, one is able to ask deeper questions to fill-in information that is necessarily missing in survey data. These spontaneous reports of users are very informative and can be used to make grounded management and service decisions.
In my experience, grounded data is more reliable for decision-making than retrospective data that may only partially represent a user's true needs. However, surveys are perceived as relatively easy to do and to analyze, while field data is viewed as too time consuming to obtain and analyze. Service decisions may in fact be more costly with incomplete data, especially in a complex area like user needs.
 Home |
 Discussion |