A Response to the weekly readings, questions about research interests 9/8

The host of readings for this week emphasized user centered theories of information use and interaction. The Mom & Me article, very clearly illustrated the concept of information seeking/use within the context of a user/situation. The article clearly pointed out issues of preferred 'ignorance' (Mom didn't care about consumer reports), decision making (Mom had her own criteria), Information/communication gap (difference between son & mom assumptions), and user perspective. These concepts were elaborated on in the other articles in more theoretical ways. Dervin & Nilan's literature review drove this point home with an extensive delineation between 'systems' oriented research and user oriented research. They outlined 3 user focused research methodologies (SenseMaking, ASK, User-Value) that all start with the user need and look back at their information seeking process. This article made me question whether or not the questions I have been asking in my research have been too systems focused and whether or not that is essentially a bad thing. As I mentioned in my opening entry, I am trying to understand/question to what extent users rely on data within a specific interface as opposed to data that is portable. Concrete examples of this are bibliographic citation management programs and data export, RSS feeds, and API interfaces to common applications (library elf for example). This semester I have begun gathering data on student use of two different interfaces for reserves at Wake. The two interfaces are 1)html based web page and 2) RSS feeds for the reserve courses. So my initial question was which 'system' would they use. Clearly this falls into what Dervin & Nilan would term a systems oriented approach. Perhaps after collecting this data I could begin to ask why they preferred one interface over another (familiarity, features, workflow). Another feature of user centered research that Dervin & Nilan mention is studying the specific or contextual user. I am concerned that the approach I have begun with again, studies only the demographic student (student in a particular course). While being able to find out if students from a particular course were more likely to use one interface over another is that getting to the 'user in context' or is it still asking a 'do they like the system' question?
I am not sure how the other two projects that I am pursuing (review of functionality of automatic metadata features in Digital library systems, and an evaluation of college student's categorization & description of primary source materials) need to be revised in light of the user/system centered question. The first seems inevitably system centered and really does not attempt to ask a deep research question. It is more a review of system functionality. Despite this, as I am attempting to conduct user surveys this semester to discover this information perhaps there are questions that I can ask - such as what kinds of AMG processing are relevant for given document collections, or what use/impact does AMG functionality have in specific usage cases. The second question (college student cataloging) seems to have more potential. I hope to study what choices the students make in their categorization decisions, how they implement structured and unstructured organization decisions, perhaps how they collectively use the cataloged information for their own retrieval, and ultimately how the description of this set of data is impacted by having 13 undergraduates perform the categorization.

I did not find the Johnson article particuarly useful, although it did detail several components of the decision making process. They tended (in my opinion) to use confrontational and biased language to discuss these issues and also emphasized a 'managerial' approach to decision making not an 'investigative' approach which recognizes the fact that users pursue information to satisfy real needs, and do not necessarily stop with only basic information, make decisions (or not ) out of fear, or control information to control others. In contrast, I felt that the Harris & Dewdney article did an excellent job of recounting these user traits and recognizing that there are situationally relevant exceptions (The user consults family members first except in sensitive cases, the user seeks convenient information except when the authority or comprehensiveness of the information is not high enough).

In general, this week was a useful re-cap of the user-oriented research that was introduced in the HCI course in the spring. Being able to come back to these ideas challenged me to re-focus and perhaps re-define my research interests.