Reality Based Interaction Workshop - CHI 2009
Presentation from Wayne Gray, RPI
Procedures we use to accomplish tasks that are important but not optimal. Use cattle roping example - horse and rider integrated into iceberg idea to lay foundation for usability versus functionality leading to a paradox of the active user.
The "Subjective Present" - fraction of a second where quite a bit is going on to connect things together. Actually observed how people scratch their head and noticed that after about 3 seconds, they either stopped or moved their hand or head. This is consistent across diverse cultures.
Embodiment Level Analysis - the time span at which cognitive, perceptive and action first occur.
Interactive Routines
Several views
Reality-based interaction - resource allocation costs expressed in currency of time - gets very finely parsed.
consider the memory as an attic where you store stuff that you aren't going to use for an extended time vs. memory as an active resource - garden hose, for the instance, use - not reflection
eye-hand span studies - making tea, music sight reading, walking on rough terrain - always about 0.5 to 2 seconds ahead
Information needed for interactive behavior should not need to be retained for more than two seconds
the interface is not a window onto the world - naive realism (Smallman 2004) - 3d good for shape recognition, 2d for judging relative position.
joint action - clark's active externalism and extended mind hypotheses - Topics in Cognitive Science
good theory - good applied problem
cogsci problems are good applied problems that have been overlooked recently within the hci community and there are strong reasons why they should be re-introduced into the community in a greater emphasis.
Presentation from Leanne Hirschfield, Tufts PhD Candidate
Emerging new interaction styles
tradeoffs - things should be based upon reality unless you make an explicit choice and know why you make that choice, focused on finding commonality of the different issues
total workload=semantic workload (task) + syntactic workload (interface)
Functional Near Infrared Spectroscopy - measuring flow of blood in pre-frontal cortex
very suitable for hci - non-invasive, high spatial resolution, easy set up, robust to noise
eeg was commonly used, fnir has some distinct advantages, only been in existence since 1990's
use benchmark workload tasks - manipulate reality (interface) and keep the task constant
trade reality for efficiency in VR world to highlight an object the user is searching for by using an arrow to reduce barriers and not inhibit the experience.
That's it for now. May try to get back to this one later as it is touching on some very interesting topics that bring cognitive science and hci together in some very useful and helpful ways.
This is Jay Steele reporting from CHI 2009 in Boston.
Procedures we use to accomplish tasks that are important but not optimal. Use cattle roping example - horse and rider integrated into iceberg idea to lay foundation for usability versus functionality leading to a paradox of the active user.
The "Subjective Present" - fraction of a second where quite a bit is going on to connect things together. Actually observed how people scratch their head and noticed that after about 3 seconds, they either stopped or moved their hand or head. This is consistent across diverse cultures.
Embodiment Level Analysis - the time span at which cognitive, perceptive and action first occur.
Interactive Routines
Several views
Brain bound - cognition "leaks" out of the mind. The pages are not a record, they are the work itself. Cognitive impartiality principle (Gray & Veskler)
Gestures - add nuances to the information we convey.
Gestures - add nuances to the information we convey.
Reality-based interaction - resource allocation costs expressed in currency of time - gets very finely parsed.
optimizes locally, not globally - design system so you have a cognitive horse and a cognitive rider
consider the memory as an attic where you store stuff that you aren't going to use for an extended time vs. memory as an active resource - garden hose, for the instance, use - not reflection
eye-hand span studies - making tea, music sight reading, walking on rough terrain - always about 0.5 to 2 seconds ahead
Does memory get out of the way after about 2 seconds to make way for new information coming in.
Information needed for interactive behavior should not need to be retained for more than two seconds
the interface is not a window onto the world - naive realism (Smallman 2004) - 3d good for shape recognition, 2d for judging relative position.
What actually sucks us in and what is useful - lots of different studies for many different industries - navy meteorologists
joint action - clark's active externalism and extended mind hypotheses - Topics in Cognitive Science
task sharing - go/no go, shared representations of tasks
good theory - good applied problem
cogsci problems are good applied problems that have been overlooked recently within the hci community and there are strong reasons why they should be re-introduced into the community in a greater emphasis.
Presentation from Leanne Hirschfield, Tufts PhD Candidate
Emerging new interaction styles
tradeoffs - things should be based upon reality unless you make an explicit choice and know why you make that choice, focused on finding commonality of the different issues
total workload=semantic workload (task) + syntactic workload (interface)
Functional Near Infrared Spectroscopy - measuring flow of blood in pre-frontal cortex
very suitable for hci - non-invasive, high spatial resolution, easy set up, robust to noise
eeg was commonly used, fnir has some distinct advantages, only been in existence since 1990's
use benchmark workload tasks - manipulate reality (interface) and keep the task constant
trade reality for efficiency in VR world to highlight an object the user is searching for by using an arrow to reduce barriers and not inhibit the experience.
That's it for now. May try to get back to this one later as it is touching on some very interesting topics that bring cognitive science and hci together in some very useful and helpful ways.
This is Jay Steele reporting from CHI 2009 in Boston.
Comments (0)
Add a Comment