#4 — Examining How Cues Affect Grounding
Our group plans to examine how conversational cues, or the lack
thereof, affect participants' behavior and perception of common
ground. Such perception can be established in three primary ways:
gestural indications, partner's activities, and salient perceptual
events (e.g., both people hear a loud noise in the next room).
To examine the nuances of perception of common ground, we will
experiment with different sets of people holding controlled
conversations in different controlled environments. For example, we
might come up with a single conversation topic, and then have
different pairs of people discuss it face to face, face to face
blindfolded, via IM, via IM with emoticons forbidden, and so on.
Blindfolds allow us to control when a participant can perceive his/her
partner's gestures and other activities. Forbidding IM emoticons will
also allow a measure of control over a participant's gestural
indications.
One possibility for experimenting with salient perceptual events is to
set up such events that only one participant can perceive and the
other is oblivious to. For instance, we may have two participants
carry on a conversation via some controlled audio chat setup, and have
a strange noise play for only one of the participants; we could then
observe how, if at all, the participants try to ground information on
what has just happened. Our primary concern about this approach,
however, is ensuring that we can set up an audio chat experiment that
we can have such control over.
By imposing different constraints on a wide variety of conversations
and participants, we will see how the medium and the constraints on
cues can affect the conversation and how the participants perceive
their common ground. In the experiments, we will use a combination of
transcripts and questionnaires answered by participants to hopefully
extract some broad observations about how different controls affect
perception of common ground, and so to better understand the roles of
different cues in communication.
~Will Fleming, representing for Stan Chen, Ron DeVera, and Josh Perlin. Word.
posted by will @ 1:38 AM
2 Comments:
I'd be very curious to see how many (or to what degree) people still gesture when they talk to another person while blindfolded. It would give a really nice indication of just how involuntary or unplanned these types of visual cues are.
As for the noise that only one participant can perceive, I feel like this type of event would be better studied in a non-experimental setting. In an experiment, I assume that people will attribute the noise to static, problems with the microphone, etc., and then quickly move on with the conversation rather than dwell on it. I'd be curious to see what you find.
I think your topic idea sounds really fascinating, although I agree with Molly in that it currently sounds a little complicated, with a lot of different variables to control. I agree that choosing one medium would probably be best (IM, FTF, etc), and then manipulating conditions within that medium to see how the grounding effects change.
I like your idea of doing something (making a noise, or whatever) and then seeing how they establish this as common ground in the various situations. I think this will have to be very noticeable/loud/etc in order for it to have any affect on the people talking; it will have to be blatantly obvious, and something that they would want to resolve or talk about with their partner. If you figure something specific out, you should be well on your way to a great experiment.
Post a Comment
<< Home