Problem-Solving Phase Transitions During Team Collaboration

Travis J. Wiltshire*, Jonathan E. Butner, Stephen M. Fiore

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

Abstract

Multiple theories of problem-solving hypothesize that there are distinct qualitative phases exhibited during effective problem-solving. However, limited research has attempted to identify when transitions between phases occur. We integrate theory on collaborative problem-solving (CPS) with dynamical systems theory suggesting that when a system is undergoing a phase transition it should exhibit a peak in entropy and that entropy levels should also relate to team performance. Communications from 40 teams that collaborated on a complex problem were coded for occurrence of problem-solving processes. We applied a sliding window entropy technique to each team's communications and specified criteria for (a) identifying data points that qualify as peaks and (b) determining which peaks were robust. We used multilevel modeling, and provide a qualitative example, to evaluate whether phases exhibit distinct distributions of communication processes. We also tested whether there was a relationship between entropy values at transition points and CPS performance. We found that a proportion of entropy peaks was robust and that the relative occurrence of communication codes varied significantly across phases. Peaks in entropy thus corresponded to qualitative shifts in teams' CPS communications, providing empirical evidence that teams exhibit phase transitions during CPS. Also, lower average levels of entropy at the phase transition points predicted better CPS performance. We specify future directions to improve understanding of phase transitions during CPS, and collaborative cognition, more broadly.

Original languageEnglish
Pages (from-to)129-167
Number of pages39
JournalCognitive Science
Volume42
Issue number1
DOIs
Publication statusPublished - Jan 2018
Externally publishedYes

Keywords

  • Problem-solving
  • Collaboration
  • Team cognition
  • Dynamical systems
  • Communication
  • NONLINEAR DYNAMICS
  • COMMUNICATION
  • COGNITION
  • MACROCOGNITION
  • COORDINATION
  • PATTERNS
  • VARIABILITY
  • PERFORMANCE
  • RECURRENCE
  • BEHAVIOR

Cite this

Wiltshire, Travis J. ; Butner, Jonathan E. ; Fiore, Stephen M. / Problem-Solving Phase Transitions During Team Collaboration. In: Cognitive Science. 2018 ; Vol. 42, No. 1. pp. 129-167.
@article{80a4c95b94c741b1be49e33708994037,
title = "Problem-Solving Phase Transitions During Team Collaboration",
abstract = "Multiple theories of problem-solving hypothesize that there are distinct qualitative phases exhibited during effective problem-solving. However, limited research has attempted to identify when transitions between phases occur. We integrate theory on collaborative problem-solving (CPS) with dynamical systems theory suggesting that when a system is undergoing a phase transition it should exhibit a peak in entropy and that entropy levels should also relate to team performance. Communications from 40 teams that collaborated on a complex problem were coded for occurrence of problem-solving processes. We applied a sliding window entropy technique to each team's communications and specified criteria for (a) identifying data points that qualify as peaks and (b) determining which peaks were robust. We used multilevel modeling, and provide a qualitative example, to evaluate whether phases exhibit distinct distributions of communication processes. We also tested whether there was a relationship between entropy values at transition points and CPS performance. We found that a proportion of entropy peaks was robust and that the relative occurrence of communication codes varied significantly across phases. Peaks in entropy thus corresponded to qualitative shifts in teams' CPS communications, providing empirical evidence that teams exhibit phase transitions during CPS. Also, lower average levels of entropy at the phase transition points predicted better CPS performance. We specify future directions to improve understanding of phase transitions during CPS, and collaborative cognition, more broadly.",
keywords = "Problem-solving, Collaboration, Team cognition, Dynamical systems, Communication, NONLINEAR DYNAMICS, COMMUNICATION, COGNITION, MACROCOGNITION, COORDINATION, PATTERNS, VARIABILITY, PERFORMANCE, RECURRENCE, BEHAVIOR",
author = "Wiltshire, {Travis J.} and Butner, {Jonathan E.} and Fiore, {Stephen M.}",
year = "2018",
month = "1",
doi = "10.1111/cogs.12482",
language = "English",
volume = "42",
pages = "129--167",
journal = "Cognitive Science",
issn = "0364-0213",
publisher = "Wiley",
number = "1",

}

Problem-Solving Phase Transitions During Team Collaboration. / Wiltshire, Travis J.; Butner, Jonathan E.; Fiore, Stephen M.

In: Cognitive Science, Vol. 42, No. 1, 01.2018, p. 129-167.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Problem-Solving Phase Transitions During Team Collaboration

AU - Wiltshire, Travis J.

AU - Butner, Jonathan E.

AU - Fiore, Stephen M.

PY - 2018/1

Y1 - 2018/1

N2 - Multiple theories of problem-solving hypothesize that there are distinct qualitative phases exhibited during effective problem-solving. However, limited research has attempted to identify when transitions between phases occur. We integrate theory on collaborative problem-solving (CPS) with dynamical systems theory suggesting that when a system is undergoing a phase transition it should exhibit a peak in entropy and that entropy levels should also relate to team performance. Communications from 40 teams that collaborated on a complex problem were coded for occurrence of problem-solving processes. We applied a sliding window entropy technique to each team's communications and specified criteria for (a) identifying data points that qualify as peaks and (b) determining which peaks were robust. We used multilevel modeling, and provide a qualitative example, to evaluate whether phases exhibit distinct distributions of communication processes. We also tested whether there was a relationship between entropy values at transition points and CPS performance. We found that a proportion of entropy peaks was robust and that the relative occurrence of communication codes varied significantly across phases. Peaks in entropy thus corresponded to qualitative shifts in teams' CPS communications, providing empirical evidence that teams exhibit phase transitions during CPS. Also, lower average levels of entropy at the phase transition points predicted better CPS performance. We specify future directions to improve understanding of phase transitions during CPS, and collaborative cognition, more broadly.

AB - Multiple theories of problem-solving hypothesize that there are distinct qualitative phases exhibited during effective problem-solving. However, limited research has attempted to identify when transitions between phases occur. We integrate theory on collaborative problem-solving (CPS) with dynamical systems theory suggesting that when a system is undergoing a phase transition it should exhibit a peak in entropy and that entropy levels should also relate to team performance. Communications from 40 teams that collaborated on a complex problem were coded for occurrence of problem-solving processes. We applied a sliding window entropy technique to each team's communications and specified criteria for (a) identifying data points that qualify as peaks and (b) determining which peaks were robust. We used multilevel modeling, and provide a qualitative example, to evaluate whether phases exhibit distinct distributions of communication processes. We also tested whether there was a relationship between entropy values at transition points and CPS performance. We found that a proportion of entropy peaks was robust and that the relative occurrence of communication codes varied significantly across phases. Peaks in entropy thus corresponded to qualitative shifts in teams' CPS communications, providing empirical evidence that teams exhibit phase transitions during CPS. Also, lower average levels of entropy at the phase transition points predicted better CPS performance. We specify future directions to improve understanding of phase transitions during CPS, and collaborative cognition, more broadly.

KW - Problem-solving

KW - Collaboration

KW - Team cognition

KW - Dynamical systems

KW - Communication

KW - NONLINEAR DYNAMICS

KW - COMMUNICATION

KW - COGNITION

KW - MACROCOGNITION

KW - COORDINATION

KW - PATTERNS

KW - VARIABILITY

KW - PERFORMANCE

KW - RECURRENCE

KW - BEHAVIOR

U2 - 10.1111/cogs.12482

DO - 10.1111/cogs.12482

M3 - Article

VL - 42

SP - 129

EP - 167

JO - Cognitive Science

JF - Cognitive Science

SN - 0364-0213

IS - 1

ER -