Analysis of Collaborative Communication for Linguistic Cues of Cognitive Load

Abstract

Objective:

Analyses of novel linguistic and grammatical features, extracted from transcribed speech of people working in a collaborative environment, were performed for cognitive load measurement.

Background:

Prior studies have attempted to assess users’ cognitive load with several measures, but most of them are intrusive and disrupt normal task flow. An effective measurement of people’s cognitive load can help improve their performance by deploying appropriate output and support strategies accordingly.

Methods:

The authors studied 33 members of bushfire management teams working collaboratively in computerized incident control rooms and involved in complex bushfire management tasks. The participants’ communication was analyzed for some novel linguistic features as potential indices of cognitive load, which included sentence length, use of agreement and disagreement phrases, and use of personal pronouns, including both singular and plural pronoun types.

Results:

Results showed users’ different linguistic and grammatical patterns with various cognitive load levels. Specifically, with high load, people spoke more and used longer sentences, used more words that indicated disagreement with other team members, and exhibited increased use of plural personal pronouns and decreased use of singular pronouns.

Conclusion:

The article provides encouraging evidence for the use of linguistic and grammatical analysis for measuring users’ cognitive load and proposes some novel features as cognitive load indices.

Application:

The proposed approach may be applied to many data-intense and safety-critical task scenarios, such as emergency management departments, for example, bushfire or traffic incident management centers; air traffic control rooms; and call centers, where speech is used as part of everyday tasks.