Group's Affetive Relevance, Affective Feedback, and Emotional Awareness (Part I)
rgonzal — Tue, 02/15/2011 - 03:31
Written by Roberto González-Ibáñez
Even though I have been writing in blogs for a while; this particular post has become more difficult of what I initially expected. In fact, I have spent the last 20 minutes in front on my computer only trying to decide what to write for the first post of our CIS Website. There are many ideas popping up in my mind right now; most of them related to the studies in which I have been involved in the past time; nevertheless, it is quite difficult to select only one of them, especially because all of them are pretty interesting to me.
Since late 2009 I started investigating about Collaborative Information Seeking (CIS), Collaborative Information Behavior (CIB), and other related research areas. This interest arose mainly from my previous experience on CSCW in which I already had worked within the context of my Master program; however, in the context of my journey as PhD student, this was facilitated by the freedom that I had in a couple of courses in the PhD program. This allowed me to explore the field of Information Science following some general directions, but more importantly, my personal motivations and research interests.
In time, I became very interested in a field that was unknown to me at that time. Perhaps the most important reason of such a growing interest was the possibility to connect my previous work, research interests, and knowledge with something within LIS.
Before enrolling in the PhD program, I did not know much about the extent of LIS; so I was quite amazed by works of authors such as Kuhlthau , Bates, Nahl, Hyldegard, Belkin, Saracevic, Wilson, Shah, among others. In the process of becoming native with the field, I found several interesting things that connected the field not only to collaboration, but also to my second research interest: affective computing.
Having these two interests in mind, a couple of class projects to do, and several inspiring articles; I came up with the idea of Groups Affective Relevance (GAR) (González-Ibáñez, R. and Shah, C. (2010a,b))… well, now I have something more defined to write about; so let me explain more about this concept:
What is Groups Affective Relevance?
Before explaining what it is, I first need to place GAR in context, so let me start by presenting some bullet points:
GAR was though for synchronous and asynchronous CIS.
GAR focus on a user-centric notion of relevance (e.g. motivational and affective relevance)
GAR can be applied in computer-mediated scenarios for collaboration as well as in non-computer-mediated settings.
With these ideas in mind, consider you have a group of people collaborating in certain project that involves seeking information; regardless the use of supporting technology or their information search strategies, I defined GAR as the overall emotional experience of each group’s member with regard to a specific information object that certain user share with the group.
What is this overall emotional experience with regard to a specific information object?
In order to answer this question let me present a complete example that will help me to address the underlying aspects of GAR.
Consider a team of 4 people working on a project proposal that requires knowledge in a variety of fields. In an initial stage team members decide to start looking for information that helps them to better justify their project. In order to evaluate the quality of the information they collect in this initial stage, the information that is found to be relevant by any member, is going to be shared with the group. By sharing the information with the group, this can be reviewed from different points of views and this may end up in agreements or disagreements among the members. What someone found to be relevant, can be supported or refuted by someone else and this lead the team to a rigorous information selection process (of course such rigurosity does apply to all scenarios of CIS, but lets keep this example to explain GAR more properly). The overall affective reactions and affective judgments with respect to a specific information object is what I call GAR; such affective reactions can be either explicit or implicit; for example someone in the group can verbally say “Hey guys, take a look to this book, I loved it. It will help us for writing the proposal”, but also there may be cases where users only express their reactions to information with gestures, physiological changes, mood, etc. The problem is how to measure GAR considering both implicit and explicit affective responses to information.
What is the importance of GAR?
GAR is presented as a natural way to measure relevance of information objects based on group consensus. Here, the word natural is used to express that relevance is captured from implicit and explicit emotional reactions. Ideally this could be supported by technology such as natural language processing, facial expression recognition, and galvanic skin response, among others .
Now, a more deep aspect of GAR is explained as follow. The interaction process among team members in the evaluation processes of information does not end in agreements or disagreements; this would be a very simplistic way of seeing this phenomenon. As human beings we are affected by interaction with others; if a person A agrees or disagrees with a person B, B is affected in multiple dimensions. One of such dimensions is the affective one.
GAR as affective feedback
Following the example of the 4 people working on a project proposal, let’s call them A, B, C, and D; if A shared I1, which is an information object that A found to be relevant. Later B, C, and D assess I1 and agree with A regarding the relevance of I1 for the project. This agreement is then considered a positive feedback for A, which may end up in better levels of confidence and happiness, among other pleasant feelings.
On the other hand, now consider B sharing I2 with the team. According to B, I2 is relevant for the project; however, after being reviewed by A, C, and D these three users completely disagree with B. This does not mean that B is wrong; however, having 3 out of 3 negative votes from the team become in a negative feedback for B, which may end up in frustration and low confidence, among other unpleasant feelings.
From GAR to emotional awareness
Collaboration as a social process involves symmetric interactions among participants and awareness. One pertinent kind of awareness related to GAR is emotional awareness (Garcia, Favela, and Machorro, 1999), which is somehow similar to empathy. In order to have a successful collaboration it is expected that members are aware of their peers in order to provide support when needed. In the examples above, if nobody is aware of what happened to B after receiving negative feedback, this may end in negative results for the entire group.
In scenarios without technology support, this emotional awareness depends mainly on the empathy of each team member. We have to add to this problem the way in which members can collaborate: face-to-face or remotely located. In face-to-face the members have several cues that may help to identify the emotional reactions of other members, but still this depends on the ability of each person to recognize and correctly interpret such cues.
[more about this soon]
- González- Ibáñez, R. and Shah, C. (2010a). Group's affective relevance: a proposal for studying affective relevance in collaborative information seeking. Poster in Proceedings of GROUP 2010. Sanibel Island, FL, USA. November 6 - 10, 2010.[pdf]
- González- Ibáñez, R. and Shah, C.(2010b). A Proposal for Measuring and Implementing Group's Affective Relevance in Collaborative Information Seeking. Poster at Workshop on Human-Computer Interaction and Information Retrieval (HCIR) 2010. New Brunswick, NJ. August 22, 2010. [pdf]
García, O., Favela, J., & Machorro, R. (1999). Emotional awareness in collaborative systems. In Proceedings of String Processing and Information Retrieval Symposium, 1999 and International Workshop on Groupware (pp. 296-303).