Jeppe Nicolaisen, from the Royal School of Library and Information Science, has an interesting forthcoming article in JASIST:
Nicolaisen, J. (in press). Compromised need and the label effect: An examination of claims and evidence Journal of the American Society for Information Science and Technology, 1-6 DOI: 10.1002/asi.21129
In a nutshell, he puts the empirical-ness of Taylor’s notion of the “compromised need” in doubt. To quickly review, according to Taylor an information need progresses through four levels when a seeker is looking for information:
- Visceral need – This is the actual, but unexpressed need for information
- Conscious need – The recognized need at a cognitive level
- Formalized need – A formal statement of the need
- Compromised need – This is the question as presented to the information system or intermediary. It called the compromised need because the inquirer must adapt the question to accommodate the available resources. This has also been called the label effect, because it has been assumed that seekers frequently fail to specify their true information needs, i.e., they use the wrong labels.
Taylor’s theory has been the inspiration for many user-centered studies in information retrieval and library science for the past three decades. It was even (part of ) the basis in my Information Search Experience model I presented at the IA Summit in 2004 in Austin. See my presentation: Information Search Experience: Emotions in Information Seeking.
But Nicolaisen finds problem after problem with studies that supposedly support the notion of the “compromised need” empirically. He is particularly critical of some works by Ingwersen, pointing out incorrect citations and interpretation of other studies. Yikes.
We have examined available studies of the compromised need / the label effect and have compared claims against evidence. The aim was to establish whether the compromised need / the label effect is a frequently occurring phenomenon or not. We found that the studies that reportedly had verified the phenomenon (Ingwersen & Kaae, ; Ingwersen, ; Belkin et al., ; Belkin, ; Nordlie, ) all suffer from technical problems that put the claim of verification in doubt. Two other studies (Lynch, ; Hauptman, ) that report low percentages of questions changing from the initial query during large-scale studies of user-librarian negotiations might indicate that users are quite often asking for precisely what they want. Although it is difficult to imagine that so many users would have accepted leaving with unanswered information needs, the fact that the librarians did not conduct in-depth interviews, and therefore may have failed to discover users’ real information needs, preclude us from making definite conclusions. However, what we can conclude is that the compromised need / the label effect is not the empirical fact that it has otherwise been claimed to be.