Recently, I conducted an exploratory study on library websites search boxes (5 Oct 2016). I wanted to know the types of search boxes deployed by libraries that are using Summon as their web scale discovery layer. I googled and discovered around 83 Summon-ed libraries from Australia, Canada, Denmark, Ireland, Saudi Arabia, New Zealand, Singapore, Sweden, UAE, United Kingdom, and the United States of America.
Here are the various types of search boxes:
There are several distinctive types:
- Simple search boxes
- Multi-Tabbed search boxes
- Search boxes with radio buttons
- Search boxes with drop down features
- Combination of multi-tabbed / radio buttons / drop down
I noted that the libraries tend to go with simple search boxes or the multi-tabbed search boxes. Out of the 83 chosen sites, 35 (42%) deployed simple search boxes while 39 (47%) deployed multi-tabbed search boxes.
I was curious whether simple or multi-tabbed would be a better choice to deploy on a library website. In this Google-era, most searchers would just enter their search phrases or keywords into the search boxes. They expect to get relevant results at the top of the list. (Bear in mind that we have a multitude of users out there: the experts, the intermediates and the novices).
Results are often determined by some of these factors (not exhaustive):
- search terms / phrases used
- metadata used to describe the library’s resources
- storage medium of these resources
- level of IT knowledge of the users
- exposure to any form of training
Coming back to the library websites, I did some sample known item searches (for example: exact title of an electronic journal) on random library websites and noted that:
- Most of them cataloged their electronic resources (e-books, e-journals etc) into their OPAC (classic catalog) even though there is evidence that they have an AZ Portal for electronic resources
Our library is ‘moving’ all our electronic resources from the classic catalog and tap into the AZ portal as the resource base. We are hoping that Summon and AZ are capable to ‘talk’ to each other. And we are implementing a feature in AZ so that these e-resources can be easily searched and located in Summon. I’m hoping that this work. Otherwise, I may have to re-think on other alternatives. Stay tune.
The best arguments for a simple search box over tabbed or any other search interface are summed up in two Nielsen Norman Group articles.
“Scoped search forces people to make a decision too soon.”
I’d argue this is the case even for just two tabs.
“Search is such a prominent part of the Web user experience that users have developed a firm mental model for how it’s supposed to work.”
Before reaching the search engine results page, this mental model consists of just two components: a box to type in; a button labeled “search”. Google’s simple search box hasn’t strayed far from this model.
If library search interface analytics indicate behaviour contrary to this firm mental model, I suggest further research to verify.
Why are tabbed search interfaces still popular? My guess is both change aversion, and infrastructure immobility rather than any other dimension described in Andy Priestner’s new model of UX adoption.
I look forward to a (predictable?) future of less tabbed and more simple search interfaces. At Deakin University Library, we’re switching ours this year. Anyone else taking the plunge?