Tuesday, March 29, 2011

Paper Reading #17: Language Complexity

Comments: 

Reference Information: 
Title: Using Language Complexity to Measure Cognitive Load for Adaptive Interaction Design
Authors: M. A. Khawaja, F. Chen, N. Marcus
Presentation: (Conference Paper) IUI'10, February 7-10, Hong Kong, China.

Summary: Cognitive Load is the mental load that a persons' memory carries when the person is performing a problem solving task. There is a limited capacity a person can maintain in order to process new information, so when a person is going through a process with large loads of information and time limitations. Such a load can be experienced by users of an interaction system. 

In this study, researches investigated patterns of speech from cognitively low and high load tasks. Their intention is to use the results obtain from the study to see how they can be used in the development of a user interface evaluation and interaction design improvements. They believe that if such a system can determine the user's cognitive load, it could adjust to the user in order to provide him or her with a better experience. 

Researches believe that the choice of words and the form of a speech is very different from written content, since there is no really much time to analyze and think about the way we want to present our ideas. Thus, it was better to carry out their studies in transcribed speeches. They obtained the data from members of bushfire management teams from Australia.

The complexity measurements they concentrated in included: Lexical Density (ratio of unique words to the total number of words), Complex Word Ratio (ratio of complex words, three syllables or more, to total number of words), Gunning Fog Index (sentence lengths and complex words), Flesch-Kincaid (estimates the number of years of education a person would require to understand text),  SMOG Grade (also focuses on the education persons require to fully comprehend text), and Lexile Level (measure of complexity).

Majority of the results obtained from the study were correct to the hypothesis they presented. With these results, they may be able to measure cognitive load of interaction systems and develop something that will aid in the interaction experience for the user.

Source
Discussion: This is the first conference paper I read that is not directly describing a technology development. In this study, researchers were interested in obtaining data that could later help in the development of a system. It was interesting to learn about the measurements that they use in order to determine the cognitive load of users. I would be really interested in reading about how the results from the study are implemented into the design of an interaction system.

When I was trying to find a relevant picture for the blog, which there weren't many, I found the following image in a scholastic website. The content of the page explained what the Lexile Level is, and how it is used with kids in order to measure their reading levels. It was interesting to read about how they are using these complexity measurements in the education field.

2 comments:

  1. This is a very interesting concept, but I think changing the UX of a system to meet their cognitive load needs, could actually create more of a cognitive load since the system is not constant. We will have to wait to see this implemented to tell for sure though.

    ReplyDelete
  2. I had never really heard of people studying language like this before. It was interesting finding out how people tend to change their behavior under stress. Hopefully this work can be developed more and help a lot in the CHI field.

    ReplyDelete