Wednesday, April 20, 2011

Paper Reading #24: Using Language Complexity to Measure Cognitive Load for Adaptive Interaction Design

Comments
Aaron Kirkes 
Felipe Othick


Reference Information
Title: Using Language Complexity to Measure Cognitive Load for Adaptive Interaction Design
Authors: M. Asif Khawaja, Fang Chen, Nadine Marcus
When/Where: IUI 2010



Summary
This paper discussed analyzing patterns of language complexity to measure cognitive load for use in interface evaluation and improving interactions. Cognitive load is the mental load on a person's memory that occurs when doing a problem-solving task due to the limitations of memory. This cognitive load can occur using interfaces due to either the task or complex design. Therefore, by analyzing users' cognitive loads, this paper hoped to improve the cognitive load problems within interfaces.
Cognitive load is due to memory limitations. Source: speakingaboutpresenting.com
The analysis was done by inspecting speech data from several bushfire management teams. The data was recorded, transcribed, cleaned, and coded for both low load and high load. By applying complexity measures to the data, it was determined that as task difficulty increases, vocabulary richness decreases and the use of complex words increases. When dealing with easy tasks, people use short, complete sentences, but do not for more difficult tasks. It is believed that this information can be applied for measuring cognitive load in other scenarios and can be used to improve interaction design.

Discussion
I found it interesting that cognitive load can be analyzed from speech data, and that this paper did so in hopes of improving interaction designs and interface evaluations. I especially liked that the paper made sure to address its faults, such as the fact that the conclusions due to the results may only apply to the particular tasks that they analyzed, namely bushfire management tasks. I think that more of these user studies should be so upfront about flaws in the data or conclusions, it makes the results more believable. Besides that, though, I think that cognitive load research could definitely result in some improvements regarding interface design and such, since it would be taking users into account during design.
Posted by at
Labels:

4 comments:

  1. VinceApril 21, 2011 at 9:39 AM

    You are right, this is actually quite an original idea. I am actually kind of shocked that sentences don't get shorter when people are under more cognitive load. I figure that people would be more likely to speak in fragments instead of using complex sentences.

    ReplyDelete
  2. Steven HennessyApril 21, 2011 at 9:43 AM

    I agree that this idea is original. I thought it was also cool they looked at speech data to analyze.

    ReplyDelete
  3. J.I.PenickApril 22, 2011 at 9:47 AM

    So the theory here is that as people get distracted, their sentences lose cohesion and they start using more complex words? That's pretty interesting, since that's not what my "shoot from the hip" expectation would have been.

    ReplyDelete
  4. Jacob LillardMay 3, 2011 at 5:56 PM

    I like that you pointed out how some papers pretend their systems are faultless. I agree that when they do this I take the paper less seriously.

    ReplyDelete

Subscribe to: Post Comments (Atom)