ArticlePDF Available

Abstract

Our objective was to apply ideas from complexity theory to derive expanded neurodynamic models of Submarine Piloting and Navigation showing how teams cognitively organize around task changes. The cognitive metric highlighted was an electroencephalography-derived measure of engagement (termed neurophysiologic synchronies of engagement) that was modeled into collective team variables showing the engagement of each of six team members as well as that of the team as a whole. We modeled the cognitive organization of teams using the information content of the neurophysiologic data streams derived from calculations of their Shannon entropy. We show that the periods of team cognitive reorganization (a) occurred as a natural product of teamwork particularly around periods of stress, (b) appeared structured around episodes of communication, (c) occurred following deliberate external perturbation to team function, and (d) were less frequent in experienced navigation teams. These periods of reorganization were lengthy, lasting up to 10 minutes. As the overall entropy levels of the neurophysiologic data stream are significantly higher for expert teams, this measure may be a useful candidate for modeling teamwork and its development over prolonged periods of training.
Nonlinear Dynamics, Psychology, and Life Sciences, Vol. 17, No. 1, pp. 67-86.
© 2013 Society for Chaos Theory in Psychology & Life Sciences
The Organizational Neurodynamics of Teams
Ronald Stevens1,The Learning Chameleon, Inc., IMMEX/UCLA, Jamie
C. Gorman,Texas Tech University, Polemnia Amazeen, Aaron Likens,
Arizona State University, and Trysha Galloway,The Learning
Chameleon, Inc
Abstract: Our objective was to apply ideas from complexity theory to derive
expanded neurodynamic models of Submarine Piloting and Navigation showing
how teams cognitively organize around task changes. The cognitive metric
highlighted was an electroencephalography-derived measure of engagement
(termed neurophysiologic synchronies of engagement) that was modeled into
collective team variables showing the engagement of each of six team members
as well as that of the team as a whole. We modeled the cognitive organization of
teams using the information content of the neurophysiologic data streams
derived from calculations of their Shannon entropy. We show that the periods of
team cognitive reorganization (a) occurred as a natural product of teamwork
particularly around periods of stress, (b) appeared structured around episodes
of communication, (c) occurred following deliberate external perturbation to
team function, and (d) were less frequent in experienced navigation teams.
These periods of reorganization were lengthy, lasting up to 10 minutes. As the
overall entropy levels of the neurophysiologic data stream are significantly
higher for expert teams, this measure may be a useful candidate for modeling
teamwork and its development over prolonged periods of training.
Key Words: teamwork, entropy, neurodynamics, EEG
INTRODUCTION
Teams have been described as complex dynamic systems that exist in a
context, develop as members interact over time, and evolve and adapt as situa-
tional demands unfold (Kozlowski & Ilgen, 2006). From the perspective of com-
plexity science, teams can be thought of as self-organized flows of information
that span biological processes and broader societal activities. As team members
interact, these often turbulent flows of information organize periodically around
a common goal only to change form again as the task and environment evolve.
In the context of the teams of which they are a part, members
continually modify their actions in response to the changing actions of others
resulting in dynamic synchronizations of information that can be observed
1 Correspondence address: Ron Stevens, The Learning Chameleon, Inc., IMMEX/UCLA,
5601 W. Slauson Ave. #272, Culver City, CA 90230, USA. E-mail: immexr@gmail.com
67
68 NDPLS, 17(1), Stevens et al.
across different systems and subsystems, including verbal (Drew, 2005),
gestural (Ashenfelter, 2007), postural (Shockley, Santana & Fowler, 2003),
functional (Gorman, Amazeen, & Cooke, 2010), physiologic (Guastello, Pincus
& Gunderson, 2006) and, more recently, neurophysiologic (Dumas, Nadal,
Soussignan, Martinerie, & Garnero, 2011; Stevens, Galloway, & Berka, 2009;
Stephens, Silbert, & Hasson, 2010). Most of these studies have consisted of two-
to-three person teams performing coordination tasks or tasks in controlled set-
tings. Our goal is to expand these ideas to larger real-world teams where the in-
formation flows are longer and expertise develops at multiple scales.
Teams, like many complex systems, are thought to operate at a level of
self-organized criticality between random and highly organized states (Bak,
Tang, & Wiesenfeld, 1987). That tenuous but significant state has also been
called the edge of chaos, a feature that allows teams to adapt to both momentary
disruptions, such as environmental perturbations, and more permanent altera-
tions, such as changes in task requirements. In this way, effective teamwork is
characterized as the continuous effort involved in stabilization of an inherently
unstable system (Gorman et al., 2010; Treffner, & Kelso, 1999). At the ‘sweet
spot’ of organization, a team demonstrates both stability and flexibility through
supportive co-regulation and adaptive team member interaction.
In keeping with the dynamics of self-organized criticality, patterns of
interaction (speech, motion, neurophysiologic changes, etc.) and activity can
change spontaneously and qualitatively with the flow of the task, and perturba-
tions to teamwork patterns are characterized by fluctuations away from and back
toward stable states across multiple levels of analysis. In a typical training se-
quence, neural events that span seconds unfold in the context of communication
events of tens of seconds that over time comprise longer, minutes-long, team
coordination events, the outcome of which influences subsequent neural events.
In that structure, we see the circular causality that is characteristic of a complex
system. When aggregated across training sessions, the tasks in which teams en-
gage provide the framework for structured formal training. The training sequen-
ce depicted in Fig. 1 spans nearly seven orders of magnitude of seconds over a
10-week course; a weakness in the literature is the lack of integrated models of
team organization that capture the linkages across these subsystems and time
scales. Such integrated models could better inform why some teams function
better than others. Are certain teams more cognitively flexible and able to more
rapidly enter and exit organized neurophysiologic states? Can these abilities be
taught, and if so, how? Longitudinal extensions of these models could be
capable of both predicting teamwork breakdowns and suggesting routes for
teams to regain their rhythm once it is lost.
Nonlinear dynamical systems (NDS) is a theoretical and methodologi-
cal approach for understanding complex systems and the linkages within and
across subsystems in a manner that deemphasizes material substrate in favor of
observed behavior patterns. NDS is a set of mathematical formalisms that can be
used to understand the time evolution of physical, behavioral, and cognitive
systems, including sudden, developmental transitions in those systems as they
evolve
.
model
s
multil
e
p
roach
gral pa
Fig. 1.
neuro
d
have
b
levels
(Steve
n
Gorm
a
stream
s
the tea
m
symbo
l
to det
e
changi
n
analys
i
chang
e
flow o
corres
p
tropy l
organi
z
In this
tions i
n
comm
u
cols fr
o
NDPLS,
1
.
One feature
s
is their app
l
e
vel systems t
h
. A second fea
r
t of the syste
m
Time scales o
f
For severa
d
ynamics of te
a
b
een developi
n
of Engagem
e
n
s et al., 201
1
a
n, 2011). Th
o
s
contain info
r
m
, and this is
l
s during diffe
r
rmine how th
o
n
g task dema
n
i
s. Based on
p
e
s in the dyna
m
f teamwork,
t
p
onding way a
n
e
vels in the d
a
z
ation of team
study, we de
s
n
the NS data
u
nication, and
n
The data set
o
m Junior Off
i
1
7(1), Or
g
ani
z
that differen
t
l
icability for
d
h
at could not
ture is the em
p
m
rather than a
f
team training
.
l
years Steve
n
a
ms in order
t
n
g models usi
n
e
nt that are t
e
1
; Stevens, G
a
o
se prior stud
i
r
mation regar
d
shown by th
e
r
ent periods o
f
o
se NS patter
n
n
ds and across
p
rio
r
results,
w
m
ics of the tas
k
t
he organizati
o
n
d the degree
o
a
ta stream, wi
t
neurophysiol
o
s
cribe team or
g
stream and b
e
n
atural and ex
t
M
P
s for these stu
d
i
cer Navigatio
n
z
ational Neur
o
t
iates dynami
c
d
escribing the
readily be c
h
p
hasis on char
a
s error.
.
n
s and collea
g
t
o detect patte
r
n
g symbolic r
e
e
rmed Neuro
p
a
lloway, Wan
g
i
es have sho
w
d
ing the curre
n
e
unequal exp
r
f
the task. A c
h
n
dynamics ca
n
different tim
e
w
e hypothesi
z
k
or encounte
r
o
n of NS dat
a
o
f organizatio
n
t
h low entrop
y
o
gic state and
h
g
anization in
t
e
gin to link th
e
t
ernal perturb
a
M
ETHOD
articipants
d
ies were coll
e
n
teams who
w
o
d
y
namics o
f
T
c
al models fr
o
behavior of
h
aracterized u
s
a
cterizing vari
a
a
gues have b
e
rns of neural
e
presentations
p
hysiologic S
y
g
, & Berka,
2
w
n that the s
y
n
t and past c
o
r
ession and or
g
h
allenge conf
r
n
be modeled
e
scales and le
v
z
ed that as te
a
r
ed perturbati
o
a
streams wou
n
could be qua
n
y
indicating a
g
h
igh entropy l
t
erms of these
e
m with team
e
a
tions in the ta
s
e
cted with IR
B
w
ere enrolled
i
T
eams
o
m conventio
n
highly compl
e
s
ing a linear
a
a
bility as an i
n
e
en studying
t
organization
a
of EEG-deri
v
y
nchronies (
N
2
011; Stevens
y
mbolic NS d
a
o
gnitive states
g
anization of
N
r
onting us no
w
in the context
v
els of teamw
o
a
ms experien
c
o
ns to the nor
m
u
ld fluctuate i
n
n
tified by the
e
g
reater degree
l
ess organizati
o
en
t
ropy fluct
u
e
xperience, te
a
s
k environmen
t
B
approved pro
t
i
n the Submar
i
69
n
al
e
x,
a
p-
n
te-
t
he
a
nd
v
ed
N
S)
&
a
ta
of
N
S
w
is
of
o
rk
c
ed
m
al
n
a
e
n-
of
o
n.
u
a-
a
m
t
.
t
o-
i
ne
70 NDPLS, 17(1), Stevens et al.
Officer Advanced Candidacy (SOAC) class at the US Navy Submarine School.
The reported data were derived from 12 Submarine Piloting and Navigation
(SPAN) simulation sessions that were selected from a total of 21 as: a) persons
in the same six crew positions were being monitored by EEG, b) the same
individuals repeated in the same positions across 2-5 training sessions over mul-
tiple days. The six members of the teams that were fitted with the EEG headsets
were the Quartermaster on Watch (QMOW), Navigator (NAV), Officer on Deck
(OOD), Assistant Navigator (ANAV), Contact Coordinator (CC), and Radar
(RAD). Additional persons participating in the SPAN who were not fitted with
the headsets were the Captain (CAPT), Fathometer reader (FATH), the Helm
(HELM), and multiple Instructors or Observers (INST).
Procedures
Submarine Piloting and Navigation sessions are required high fidelity
navigation training tasks, and each session contains three segments, beginning
with a Briefing in which the overall goals of the mission are presented. The Sce-
nario is a dynamically evolving task containing both easily-identified and less
well-defined teamwork processes. The Debriefing following the Scenario is the
most structured part of the training; it is a topical discussion of what worked and
what other options may have been available along with long- and short-term
lessons. One regularly-occurring process during the Scenario is the periodic up-
dating of the ship’s position, termed ‘Rounds’. In taking Rounds, three naviga-
tion points are chosen, and the bearing of each from the boat is measured and
plotted on a chart. This process occurs every three minutes with a countdown
from the one-minute mark, where the Recorder logs the data (Fig. 2A). A
sample navigation task is diagrammed in Fig. 2B: The submarine (whose route
is indicated by the black circles with time offsets) was being steered northward
(up) and its position is identified by number at different times (epochs or
seconds). The submarine encountered an outbound ship (~ epoch 850), an in-
bound merchant (~ epoch 2100), and an outbound merchant (~ epoch 2100),
each requiring changes in course or speed to avoid collision. In Fig. 2A, the top
team showed a regular progression of the five-step sequence, being irregular at
only two points (gray). The second team showed a more disrupted Rounds
process. Quantitative internal and external outcome measures are generally not
available from SPAN as formative and summative feedback is a group process
in the style of Total Quality Management (Ahire, 1997). We have attempted to
develop an internally-derived outcome measure from the frequency or
completeness of the Rounds sequences.
The regularity of the Rounds countdown, along with possible
deviations, was obtained from the speech of the Recorder. When only three (or
fewer) steps of the Rounds sequence were completed, or when an entire Rounds
sequence was missed, it often indicates a team that is experiencing difficulty.
The outcome measure is simply the percentage of completed Rounds sequences.
For in
s
possi
bl
T4S2
o
Fig. 2.
SPAN
subma
r
by the
contai
n
softwa
r
NDPLS,
1
s
tance, during
l
e rounds se
q
u
o
nly contained
Components
o
teams. B: T
h
r
ine and other
b
lack circles b
e
The Adva
n
n
s an easily-
a
r
e designed t
o
1
7(1), Or
g
ani
z
the SPAN pe
r
u
ences were
c
eight complet
e
o
f SPAN tasks.
h
e numbers
o
traffic during t
h
e
ginning at 59
0
M
n
ced Brain
M
a
pplied wirel
e
o
identify and
z
ational Neur
o
r
formance E2
S
c
ompleted, w
h
e
d of 17 possi
b
A: The seque
n
o
n the tracks
h
e simulation;
0
seconds.
M
easures
M
onitoring, In
c
e
ss EEG sys
t
eliminate mul
t
o
d
y
namics o
f
T
S
1 in Fig. 2A
h
ereas the SP
A
b
le Rounds se
q
nce of Round
s
indicate the
the submarin
e
c
. (ABM),
B
tem that inc
l
t
iple sources
o
T
eams
12 of 15 (80
%
A
N performa
n
q
uences.
s
is shown for t
w
position of
t
e
’s track is sho
w
B
-Alert® syst
e
l
udes intellig
e
o
f biological
a
71
%
)
n
ce
w
o
t
he
w
n
e
m
e
nt
a
nd
72 NDPLS, 17(1), Stevens et al.
environmental contamination and allow real-time classification of cognitive
state changes even in challenging environments. The nine-channel wireless
headset includes sensor site locations F3, F4, C3, C4, P3, P4, Fz, Cz, and POz in
a monopolar configuration referenced to linked mastoids. ABM B-Alert® soft-
ware acquires the data and quantifies alertness, engagement, and mental work-
load in real-time using proprietary software (Berka, Levendowski, Cvetinovic,
Petrovic, & Davis, 2004). Data processing begins with the eye-blink decontami-
nated EEG files that contain second-by-second calculations of the probabilities
of High EEG-Engagement (EEG-E) and High EEG-Workload (EEG-WL).
Simple baseline tasks are used to fit the EEG classification algorithms to the
individual so that the cognitive state models can then be applied to increasingly
complex task environments. The EEG-E metric is an approximation of the
multiple ways in which the term Cognitive Engagement has been reported in the
literature. For instance, it has been used to describe the amount of cognitive
processing that a learner applies to a subject (Howard, 1996) or as something
that has to be broken during a task so that a learner can reflect on his or her
actions (Roberts &Young, 2008). It shares similarities with alertness or attention
and can be visual or auditory. It is analogous to the EEG-rhythm-based attention
measures that are often associated with alpha power dynamics (Jung, Makeig,
Stensmo, & Sejnowski 1997; Kelly, Docktree, Reily, & Robertson, 2003;
Huang, Jung, & Maekig, 2007). Operationally, precise cognitive terms will be
difficult to associate with EEG-derived measures of cognition in the context of
teamwork, and functional associations will need to be derived empirically.
Analytic Procedures
Neurophysiologic methods can extend the use of speech for modeling
team dynamics by providing “in the head” measures of team dynamics (Warner,
Letsky, & Cowan 2005). As team members interact and perform their duties,
each would be expected to exhibit varying degrees of cognitive states such as
attention, workload, or engagement. We assume that the levels and patterns of
variability of these components across team members reflect aspects of team
cognition. Rather than focus on neurophysiologic markers, such as P300 or
N400 that rapidly appear and disappear in response to many stimuli, we have
used EEG-Eor EEG-WL which tend to persist longer across teams.
Neurophysiologic synchrony models were developed by first aggregat-
ing the second-by-second EEG-E levels from each team member into a six-unit
vector. We used an unsupervised artificial neural network (ANN) with a linear,
competitive architecture to extract from these vectors collective team variables
termed neurophysiologic synchronies of engagement (NS_E) that showed the
engagement of each of six team members as well as of the team as a whole
(Stevens, Galloway, Wang, & Berka, 2011). ANN classification of these
second-by-second vectors created a symbolic state space that showed the
possible combinations of EEG-E across members of the team. Figure 3 shows
three symbols that illustrate the diversity of EEG-E levels across team members.
They are samples from the 25 symbols in Fig. 4A.
ment c
occupi
e
thoug
h
time is
of tra
n
attract
o
Fig. 3.
Each
b
team
m
Fig. 4.
set of
s
twenty
-
neuro
p
resulti
n
assum
p
among
of tea
m
most s
i
config
u
matrix
distrib
u
time t
a
numbe
r
shown
a pro
m
the NS
the str
u
NDPLS,
1
Established
an be describ
e
e
s in the stat
e
h
t of as the sta
t
indicated by
t
n
sitions, over
o
r.
Three NS sy
m
b
ar in the diff
e
m
ember.
Developing th
s
tate variables
-
five NS sta
t
hysiologic sy
n
n
g from the ran
For ANN tr
a
p
tion that mos
t
individual te
a
m
re-organizat
i
i
milar states
w
u
ration should
if most transi
t
u
ted. Transitio
a
gainst the N
S
r
s at each tran
by the heat m
a
m
inent diagona
l
state were s
m
u
cture, or info
r
1
7(1), Or
g
ani
z
dynamical m
o
e
d using a set
o
e
space. In o
u
t
e variables, a
n
t
he pattern of
N
both shorter
m
bols resultin
g
e
rent symbols
e dynamics of
and their tran
s
t
e variables
n
chronies stat
e
domization of
t
a
ining, we us
e
t
secon
d
-by-se
c
a
m members a
n
i
on. The line
a
w
ere proximal
result in a di
a
t
ions were loc
n
matrices pl
o
symbol num
b
sition are sum
m
a
ps. The trans
i
l
indicating th
a
m
all (Fig. 4B).
W
r
mation, in the
z
ational Neur
o
o
dels of agen
t
s
o
f state variab
l
u
r system, the
n
d the positio
n
N
S state trans
i
and longer ti
m
g
from artifici
a
represents th
e
neurophysiol
o
s
itions from t t
o
can be trac
k
e
transition m
t
he NS data st
r
e
d a linear arc
h
c
ond state tra
n
n
d that larger
t
a
r architecture
and that diff
e
a
gonal line in
a
l and in a di
s
o
t the NS sym
b
b
er expressed t
h
m
ed over the
p
i
tion matrix o
f
a
t many of th
e
W
hen the NS_
E
NS data strea
m
o
d
y
namics o
f
T
s
interacting
w
l
es and the po
s
different NS
n
of the syste
m
i
tions. We con
me steps, to
a
l neural netw
o
e
EEG-E acti
v
o
gic synchrony
o
t+1. A: The a
c
k
ed over ti
m
atrix. C: The
r
eam in B.
h
itecture of no
d
n
sitions would
t
eam shifts wo
of the ANN
e
rences were
m
a secon
d
-by-
s
s
persed map i
f
b
ol number b
e
t
he next secon
d
p
erformance,
a
f
the NS_E da
t
e
second-by-s
e
E data stream
m
was lost (Fi
g
T
eams
w
ith the envir
o
sition the syst
e
symbols can
m
at any point
sider that patt
e
be a dynami
c
o
rk classificati
o
v
ity levels of
o
attractors by
t
c
tivity level of
t
m
e using a
(
transition ma
t
des on the ini
t
be local chan
g
o
uld be indicat
i
ensured that
t
m
ore distal. T
h
s
econd transit
i
f
they were m
o
e
ing expresse
d
d
(i.e. t + 1).
T
a
nd the totals
a
t
a stream sho
w
e
cond changes
was randomiz
e
g
. 4C).
73
o
n-
e
m
be
of
e
rn
c
al
o
n.
o
ne
t
he
t
he
(
B)
t
rix
t
ial
g
es
i
ve
t
he
h
is
i
on
o
re
d
at
T
he
a
re
w
ed
in
e
d,
74
stream
.
extract
i
task, t
e
ternal
t
Fig. 5.
the tra
n
perfor
m
data fo
deter
m
the tas
k
SPAN
Briefi
n
expres
s
more
r
segme
n
distrib
u
transit
i
The data in
.
The goal wa
s
i
ng the infor
m
e
am performan
t
ask perturbati
o
Sub-task distr
n
sition matrix
a
m
ance by a S
O
r the three ma
j
Captur
i
We used th
e
m
ine how team
k
. Figure 5 sh
o
performance
n
g, Scenario, a
s
ion was seen
r
estricted in
N
n
ts showing
u
tions. The m
o
ons suggestin
g
NDPLS,
1
R
Fig. 4 indica
t
s
to build an
o
m
ation containe
ce and experti
s
o
ns.
i
butions of NS
a
nd expressio
n
O
AC team. Th
e
or segments
o
i
n
g
Tas
k
-Ind
u
e
three-
p
art st
r
organization
a
o
ws the NS sy
m
that had bee
n
n
d Debriefing
with the enti
r
N
S symbol ex
p
more com
p
o
st frequent
N
g
the persisten
c
1
7(1),
S
tevens
R
ESULTS
t
ed there was
o
rganizational
d in the NS d
a
s
e, team com
m
symbols and
t
n
of the twenty
-
e
matrices and
o
f the task.
u
ced Shifts in
r
ucture of the
S
a
t a neurodyn
a
m
bol frequenc
i
n
decomposed
segments. T
h
r
e SPAN sess
i
p
ression, wit
h
p
lementary r
a
N
S symbols w
e
c
e of symbol
e
et al.
structure in t
h
model of nav
i
a
ta stream and
m
unication, an
d
t
ransitions. Th
e
-five NS symb
histograms b
e
NS Distribut
i
S
PAN task in
a
mics level w
i
es and transiti
into periods
h
e greatest het
e
i
on. Each SP
A
h
the Scenari
o
a
ther than
o
e
re also highl
i
e
xpression. Fr
o
h
e NS_E sym
b
igation teams
relating it to
t
d
internal and
e
e
top level sho
ols for the SP
A
e
low show sim
i
i
ons
the first study
as influenced
i
on matrices f
o
represen
t
ing
t
e
rogeneity in
N
A
N segment
w
o
and Debrief
i
o
verlapping
N
i
ghted in the
N
o
m the NS dis
t
b
ol
by
t
he
e
x-
ws
A
N
i
lar
to
by
o
r a
t
he
N
S
w
as
i
ng
N
S
N
S
t
ri-
bution
s
mem
be
ring t
o
N
S_E
where
11). O
v
tions o
f
Fig. 6.
period
SPAN
agonal
chang
e
al que
s
they p
e
from t
i
segme
n
NDPLS,
1
s
, there were
fe
e
rs simultaneo
u
o
the NS sym
b
symbols 14,
1
the majority
o
v
erall, the pat
t
f
the team occ
u
NS_E transit
of a SPAN
D
p
erformances
c
Dynami
c
The fact th
a
suggests NS
e
from one per
s
s
tions include
h
e
rsist. Figure
6
i
me t (X axi
s
n
t.
1
7(1), Or
g
ani
z
e
w periods in
u
sly and persi
s
b
ol map; that
1
5, 21, and 2
4
o
f the team m
e
t
erns of NS ex
p
u
r with chang
e
ion matrix sa
m
D
ebriefing. Se
c
c
an be found
a
c
s of NS Attr
a
a
t the most fre
q
state persiste
n
s
istent state to
h
ow rapidly th
e
6
tracks the
N
s
) to time t
+
z
ational Neur
o
a
ny of the Sc
e
s
tently had hi
g
condition wo
4
. Instead, the
e
mbers had lo
w
p
ression sugg
e
e
s in task dem
a
m
pled at diffe
r
c
ond-by-secon
a
t www.teamn
e
a
ctor Formati
o
q
uent NS in e
a
n
ce but offers
another. From
e
se states dev
e
N
S_E (state v
a
+
1 (Y axis)
o
d
y
namics o
f
T
e
nario segmen
t
g
h levels of e
n
o
uld have bee
n
dominant sy
m
w
E (i.e., NS
_
e
st that qualit
a
a
nds.
r
ent points ov
e
n
d dynamics
o
e
urodynamics.
c
o
n and Dispe
r
a
ch task segm
e
little about h
o
m
a dynamics p
e
e
lop and dispe
r
a
riable) transit
i
during one S
P
T
eams
t
s where all te
a
n
gagement; re
f
n
represented
m
bols were th
o
_
E symbols 10
a
tive re-organi
z
e
r a 584 sec
o
o
f this and ot
h
c
om.
r
sion
e
nt lie on the
o
w the activit
e
rspective, nat
u
r
se and how l
o
i
ons of the te
a
P
AN Debrief
i
75
a
m
f
e
r
-
by
o
se
&
z
a-
o
nd
h
er
di-
ies
ur
-
o
ng
a
m
i
ng
76 NDPLS, 17(1), Stevens et al.
The transition matrices in Fig. 6 are sequential snapshots of the system
at times following the first frame when an attractor region around NS symbols
1-4 began to form. As this activity increased, the smaller transition regions
around NS symbols 20 and 11 began to disperse, and by 96 seconds the activity
in the region of NS symbols 1-4 dominated. This area remained stable for the
next two minutes (until 320 seconds) and then began to disperse with the
appearance of new transitions around from NS 1 to NS 20. This area was stable
for the next two minutes, and there were reciprocal from -> to transitions across
NS symbols 1 and 20. Two possible interpretations are: (a) that this is a periodic
attractor or (b) that the pattern represents a sequence of attractors that that form
or dissolve with changes in task demands. After approximately two more
minutes (at 584 seconds), the activity around NS 20 dominated. This sequence
of attractor formation is informative because whereas most NS transitions are
local, as indicated by the diagonals in Figs. 4 and 5, phase transitions often
begin by temporary transitions far from the diagonal of the transition matrix.
Though a symbolic representation of the state of the team is useful for
characterizing team neurodynamics, it is not the best tool for quantifying team
neurodynamics. Although there are methods for the quantitative representation
of symbols (Daw, Finney & Tracey, 2003), we chose to perform a moving
average window approach to derive numeric estimates of Shannon entropy of
the NS symbol stream. Shannon entropy is the informational content of the sym-
bol stream measured by the number of binary decisions (calculated in bits) re-
quired to represent the symbol stream at a given point in time (Shannon &
Weaver, 1949). The NS entropy measure captures the distribution of activity
across the state space. In terms of team cognition, low entropy may be inter-
preted as a highly-ordered team neurophysiologic state, whereas high entropy
would correspond to a more random mix of team neurophysiologic states. The
maximum entropy for 25 randomly-distributed NS symbols is log2 (25) = 4.64.
In comparison, an entropy value of 3.60 would result if roughly half (12) of the
NS symbols were randomly expressed. To develop an entropy profile over a
SPAN session, the NS Shannon entropy was calculated at each epoch using a
sliding window of the values from the prior 100 seconds. Windowing over
longer periods decreased the resolution of entropy changes, whereas smaller
windows (e.g. 30 seconds) increased the potential for false positives. An inter-
esting feature of the attractor sequence in Fig. 6 was the changing levels of en-
tropy in the NS data stream, which are shown by the bar to the right of each
frame. Periods of low entropy were associated with changes in the shape of the
attractor. Our work represents a preliminary step in the use of entropy and its
dynamics to understand the real-time organization of team cognition. More
information is needed on what drives teams to these areas of high organization,
and whether this organization is beneficial to the team.
NS_E Dynamics Are Not Uniform
The previous neurodynamic models are expanded in Fig. 7 for another
SPAN team session. This sequence of figures illustrates the transformation of
seque
n
zation.
bols.
F
tropy
f
expres
s
larly a
t
13-18
Debrie
Fig. 7
.
secon
d
NS_E
s
entrop
y
the 62
5
were u
s
fluctu
a
smalle
r
We ar
e
other
d
Gallo
w
E
n
time t
h
NDPLS,
1
n
ces of
N
S sy
m
Figure 7A sh
o
F
igures 7B an
d
fl
uctuations of
s
ions of the N
S
t
the task junc
t
were poorly
f
ing.
.
Multiple rep
r
d
expression
o
s
how the NS
_
y
profile (C). D
5
potential (i.
e
s
ed by the tea
m
The variati
o
a
tions coverin
g
r
and shorter
f
e
currently ex
d
ynamical an
a
w
ay, in prepara
t
n
trop
y
Fluctu
a
Patterns of
h
an it takes to
u
1
7(1), Or
g
ani
z
m
bols into a q
u
o
ws the secon
d
d
7C show the
the NS data st
r
S
symbols wer
e
t
ions (indicate
d
expressed d
u
r
esentations o
f
o
f individual N
S
_
E symbols be
i
uring periods
o
e
. from 25 sy
m
m
during a 10
0
o
ns in the NS
_
g
minutes.
Ne
f
luctuations (i.
plo
r
ing the fr
a
lysis techniq
u
t
ion).
a
tions can be
A
NS_E entrop
y
u
tter a questio
n
z
ational Neur
o
u
antitative me
a
d
-by-second e
x
attractor stat
e
r
ea
m
. As with
e
not uniform
b
d
by the arro
w
u
ring the Sc
e
f
NS_E neur
o
S
_E symbols.
i
ng expressed
o
f low entropy
m
bols to 25 s
y
0
second wind
o
_
E entropy lev
e
sted within
t
e. the NS ent
r
a
ctal nature o
u
es (Likens,
A
A
ssociated wi
t
y
fluctuation
c
n
or sentence,
s
o
d
y
namics o
f
T
a
sure of the da
t
x
pression of t
h
e
s associated
w
most SPAN
p
b
ut changed o
v
w
s. For instanc
e
e
nario but d
o
o
dynamics. A:
B: The transi
at the region
s
(~epochs 190
y
mbols) NS s
y
o
w.
els were com
p
t
hese larger
fl
r
opy streams
a
o
f these entro
p
A
mazeen, Gor
m
t
h Conversat
i
c
an last for c
o
s
uggesting tha
t
T
eams
t
a stream orga
n
h
e 25 NS_E sy
m
w
ith different
e
p
erformances,
t
v
er time, parti
c
e
, NS_E symb
o
o
minated in
t
The second-
b
tion matrices
s
indicated in
t
0 & 2400) fe
w
y
mbol transiti
o
p
lex, with lon
g
fl
uctuations w
e
a
ppeared fract
a
p
y streams us
i
m
an, Stevens,
i
on Episodes
o
nsiderable m
o
t
if an associat
i
77
n
i-
m
-
e
n-
t
he
c
u-
o
ls
t
he
b
y-
for
t
he
w
of
o
ns
g
er
e
re
a
l).
i
ng
&
o
re
i
on
78
exists
b
higher
-
to the
constr
u
episod
e
to be
c
that ar
e
satisfa
c
as a di
s
Debrie
discus
s
4800 t
o
the su
b
OOD
a
the N
S
rapidl
y
was r
e
observ
e
causal
i
Fig. 8.
sodes
b
bar is
entrop
y
NS
_
p
ause
d
cerns
a
in Fig.
while
t
observ
e
b
etween NS
e
-
level discour
s
“episodes” d
e
u
cted sequenc
e
e
, it is based
o
c
onsistent with
e
linked to th
o
c
tory to all, or
s
cussion aroun
Figure 8 sh
o
f
ing segment
s
ion topics: o
n
o
5285. In the
b
marine devi
a
a
sked the tea
m
S
_E entropy
s
y
increased an
d
e
ached. Impo
r
e
d in the ab
o
ty in that we c
a
NS entropy o
r
b
ar shows the
color coded t
o
y
variability sh
o
_
E Entrop
y
F
There were
d
while the Ca
p
a
nd recommen
d
9. Coincide
n
t
he team re-o
r
e
d a rapid shif
t
NDPLS,
1
e
xpression an
d
s
e units. Thes
e
e
scribed by S
e
s of behavi
o
o
n the premise
their percepti
o
se of others.
it might conti
n
d a central the
m
o
ws detailed
N
of one SP
A
n
e from epoc
h
first segment,
a
ted around a
m
if they unde
r
s
teadily dropp
e
d
again slowly
d
r
tantly, these
o
ve studies w
e
a
n only infer
w
r
ganizes aroun
major discuss
o
periods wh
e
o
ws the entrop
y
F
luctuations
O
two instance
s
p
tain or Navig
a
d
ations. The
N
n
t with the pa
u
r
ganized itsel
f,
t
up to the pri
o
1
7(1),
S
tevens
d
speech, the
n
e
highe
r
-level
d
alem (2011).
o
r. When co
n
that individu
a
ons and then
e
The episode
m
n
ue into anoth
e
m
e or topic.
N
S entropy ma
p
A
N performa
n
h
s 4171 to 48
0
the team eng
a
merchant, an
d
r
stood his ove
r
e
d until clos
u
d
eclined as cl
o
fluctuations
i
e
re natural pr
o
w
hat induced t
h
d conversatio
n
ion episodes
o
e
n there were
y
profile.
O
ccur Around
s
when the S
P
a
tor addresse
d
N
S_E profile f
o
u
se was a gra
d
f,
and at the c
r, less-organiz
et al.
n
it may be
o
d
iscourse unit
s
Episodes co
n
n
versation is
a
ls initially co
n
e
volve these
m
m
ay evolve u
n
e
r episode. It
c
p
ping of episo
d
n
ce. There
w
0
0 and a sec
o
a
ged in a disc
u
d
, in the sec
o
r
all plan. Duri
n
u
re was reach
o
sure on the se
i
n entropy a
n
o
ducts of tea
m
h
em.
n
al episodes o
r
o
f the Debriefi
n
different spe
a
Perturbation
s
PAN Scenari
o
d
the navigatio
n
o
r one of thes
e
dual decline i
onclusion of
t
z
ed team state.
o
rganized aro
u
s
m
ay be sim
i
n
sist of mutu
a
described as
n
struct messa
g
m
essages in w
a
n
til it is mutu
a
c
an be thought
d
e shifting in
t
w
ere two ma
j
o
nd from epo
c
u
ssion about
w
o
nd segment,
t
n
g the first to
p
ed. The entr
o
cond major to
p
n
d the attract
o
m
work and l
a
r
topics. The
E
n
g. The Speak
e
a
kers. The NS
_
s to the Tas
k
o
was extern
a
n
team with c
o
e
events is sho
w
i
n NS_E entr
o
t
he discussion
u
nd
i
lar
a
lly
an
g
es
a
ys
a
lly
of
t
he
j
or
c
hs
w
hy
t
he
p
ic
o
py
p
ic
o
rs
a
ck
E
pi-
e
rs
_
E
a
lly
o
n-
w
n
o
py
is
Fig. 9.
period
organi
z
togeth
e
team
p
N
S_E
Round
s
The e
x
step R
o
was a
smoot
h
irregul
a
Round
s
often
i
equip
m
had lo
w
more r
regula
r
levels
o
Experi
e
compa
r
entrop
y
NDPLS,
1
Perturbation
o
highlighted, t
h
z
ed than after
t
Linkin
g
NS
_
Many of t
h
e
r in Fig. 10,
w
p
erformance.
F
entropy, a p
r
s
performance
x
pert team ses
s
o
unds countd
o
more pattern
e
h
NS_E entr
o
a
r for tea
m
S
O
s
sequences,
w
i
ndicate stres
s
m
ent failures (
S
w
er overall N
estricted trans
i
r
ity of Round
s
o
f NS_E entro
p
These find
i
e
nced (SUB,
n
r
isons were
m
y
levels whe
r
1
7(1), Or
g
ani
z
o
f the SPAN t
h
e simulation
w
t
he pause.
_
E Entrop
y
Fl
h
e findings
d
w
hich provide
s
F
igure 10 sho
w
r
ofile of the
e
metric for a
m
s
ion in Fig. 10
A
o
wns and also
h
e
d backgroun
d
o
py profile.
T
O
AC 2 where
w
ere missed a
s
ful condition
S
tevens, Gallo
S_E ent
r
o
p
y
l
i
tion matrix.
T
s
taking, whic
h
p
y (Fig. 10B).
i
ngs were ex
p
n
=6) and SO
A
m
ade (Table 1
r
e Experience
d
z
ational Neur
o
ask induces t
e
w
as in pause
u
ctuations wi
t
d
escribed in
p
s
a framework
w
s the NS_E
e
ntropy fluctu
a
m
ore experien
c
A
showed mo
s
h
ad the highe
s
d
in the tran
s
T
he Rounds
s
individual st
e
s
indicated b
y
s like makin
g
way, Wang
&
l
evels with m
o
T
here was a p
o
h
is an intern
a
p
lored using
A
C (n=6) nav
): The first
w
d
teams had
o
d
y
namics o
f
T
e
am reorganiz
a
and the attra
t
h Team Perf
o
p
revious secti
o
for linking N
S
transition m
a
ations, and t
h
ced (SUB) an
d
stly regular a
n
s
t overall NS_
E
s
ition matrix
s
equence patt
e
e
ps, and occas
i
y
the gray bo
x
g
a turn, av
o
&
Berka, 2011
)
o
re fluctuatio
n
o
sitive correla
t
a
l performanc
e
NS_E comp
a
v
igation teams
.
w
as across th
e
significan
t
ly
h
T
eams
a
tion. During
t
ctors were m
o
fo
rmance
o
ns are brou
g
S
_E entropy
w
a
trix, the ove
r
h
e output of
t
d
a SOAC te
a
n
d complete fi
v
E
entropy. Th
e
and a relativ
e
e
rns were m
o
ionally compl
e
x
es. Irregularit
o
iding traffic
)
. This team a
l
n
s and showe
d
t
ion between
t
e
metric, and
t
a
risons betw
e
.
Three differ
e
e
average NS
_
h
igher levels
79
t
he
o
re
g
ht
w
ith
r
all
t
he
a
m.
v
e-
e
re
e
ly
o
re
e
te
ies
or
l
so
d
a
t
he
t
he
e
en
e
nt
_
E
of
80
entrop
y
repres
e
sizes
o
The i
d
organi
z
showe
d
entaile
d
structu
r
that a
s
1, the
S
the SU
Fig. 1
0
Round
s
above
for the
the ta
k
segme
y
. The second
e
nted in the tr
a
o
f the transitio
n
d
ea was that
s
z
ed performa
n
d
a more high
l
d
recurrence
r
e in noisy, c
o
s
ystem revisits
S
UB and SO
A
B teams show
i
0
. Entropy fluc
t
s
for a repre
s
the NS_E ent
r
Scenario seg
m
k
ing of Round
s
n
ts of three S
U
NDPLS,
1
comparison i
n
a
nsition matric
e
n
matrices of
t
s
ince PNG fi
l
n
ces will hav
e
l
y organized s
t
quantificatio
n
o
upled dynam
i
similar states
A
C teams wer
e
i
ng fewer recu
r
t
uations and s
e
s
entative expe
r
r
opy profiles.
T
m
ent. Figure 1
s
, against the
o
U
B and three S
1
7(1),
S
tevens
n
directly meas
u
e
s. This was p
e
t
he different
S
l
es provide l
o
e
the smallest
t
ate by the S
O
n
analysis, a
i
c systems by
(Webber & Z
b
significantly
d
r
rences than t
h
e
quencing of
t
r
ienced (SUB
)
T
o the right ar
e
0B plots the
o
o
verall NS_E
e
OAC SPAN te
a
et al.
u
red the degre
e
e
rformed by c
o
S
UB and SO
A
o
ssless compr
e
file size. Th
i
O
AC teams. T
h
tool for ext
r
quantifying t
h
b
ilut, 2005). A
s
different by t
h
h
e SOAC team
t
he Rounds. T
)
and SOAC
t
e
the overall tr
a
o
utput of a per
f
e
ntropy levels
a
ms.
e
of organizat
i
o
mparing the
f
A
C performanc
e
ssion, the m
o
i
s approach a
l
h
e third appro
a
r
acting tempo
h
e points in ti
m
s shown in Ta
b
h
is measure,
w
s.
he sequences
t
eam are plot
t
ansition matri
c
r
formance met
r
for the Scen
a
i
on
f
ile
es.
o
st
l
so
a
ch
o
ral
m
e
b
le
w
ith
of
t
ed
c
es
r
ic,
a
rio
NDPLS, 17(1), Organizational Neurodynamics of Teams 81
Table 1. Comparisons between Experienced (SUB, n=6) and SOAC (n=6)
navigation teams.
NS_E Entropy Transition Map
Size (bytes) Percent
Recurrences
Expert 4.22 ± 0.01 15,072 ± 2,232 1.05 ± 0.62
SOAC teams 4.08 ± 0.12 12,068 ± 2,807 3.2 ± 1.60
Significance p < 0.001
Kruskal-Wallis
test
p < 0.04
Wilcoxen p <0.007
t-Test
(independent)
These results indicate that, on the average, experienced teams have
fewer periods of decreased NS entropy, or the decreases have a shorter period or
amplitude, suggesting a less organized state than the SOAC teams.
DISCUSSION
The results presented in this paper show that the NS symbol streams
contain multiple levels of structure that relate to the functioning of SPAN teams.
At the simplest level, the NS_E entropy values, and presumably the sequence of
NS_E symbols, are not random but have a structure. Part of the structure is
imposed by the modeling system, where the linear architecture of the
unsupervised ANN is designed so that similar symbols are located nearby and
more different symbols are located further away. We took advantage of this
architecture to show that many of the second-to-second changes in the EEG-E
levels of the team occur in local neighborhoods. This does not mean that the
NS_E transitions off the diagonal are noise. Instead, they may signal the onset of
a significant shift across the state space. The dynamics of these shifts were
interesting because they often exhibited reciprocal transitions across two NS
symbols resulting in a four-point transition matrix pattern as illustrated in the
320 second and 432 second panels of Fig. 6. Few of these off-axis transitions
persisted longer than several minutes, and the system eventually stabilized on or
near diagonal transition, which would seem to be the attractors of the system.
This is further suggested by the association of different attractors with different
segments of the task.
A second level of structure was the fluctuations in the NS_E entropy
stream. The periods of team cognitive re-organization identified by entropy
fluctuations: (a) occurred as a natural product of SPAN teamwork (Figs. 7 and
10), (b) appeared linked with episodes of communication (Fig. 8), and (c) were
associated with external perturbations to teamwork (Fig. 9). Evidence is
beginning to accumulate suggesting that periods of intensity or stress contribute
to the natural decreases in NS_E entropy. These decreases indicate not only a
change in organization but increased organization. There is a substantial
psychology literature on the importance of conflict on the synchronization of
group communication and interactions (Pincus, 2009). Most relevant for this
82 NDPLS, 17(1), Stevens et al.
study are the physiological synchronizations in personal relationships
characterized by conflict. Such conflict causes structural changes in
interpersonal dynamics by shifting the individuals and groups into a more
organized (i.e. rigid) state of thinking and acting. This parallels our findings of
periods of increased team organization being associated with increased team
stress due to visibility, the number of contacts in the vicinity, restricted
maneuverability, etc., (Stevens et al., 2011). Though the SUB navigation teams
encountered simulation events similar to those of SOAC teams, their increased
training or experience did not cause interruptions or restrictions to the flow of
cognitive information among the team members.
The patterns of neurophysiolgic organization could be lengthy, lasting
up to 10 minutes, and were often more associated with communication episodes
than shorter ‘thought units’ including sentences, utterances, or who was speak-
ing. In the Debriefing segments, where speech is synchronous and most highly
structured, there are intriguing associations between NS_E entropy and episodes
of conversation that need to be further explored. These studies, and others being
performed with a simpler map tracing task, suggest that the NS organizations are
not only speaker or listener responses (Stephens et al., 2010) but also reflect
longer periods of deliberation by the team.
In a broad sense, we view teams as real-time dynamical systems that
must continuously adapt to changes in task requirements and unpredictable per-
turbations to remain effective. Of course, some teams are better at this than
others, and metrics based on communication analysis and other aspects of team
performance have been developed to detect subtle differences in team effective-
ness. Importantly, the team neurosynchrony studies presented in this paper re-
vealed expert or novice differences, which typically manifest themselves over
relatively long time scales of team development.
We have integrated these data with performances from other teams that
we have studied into a model linking NS_E entropy and state transitions with
experience and perhaps the development of expertise (Fig. 11). The cognitive
organization axis reflects the overall entropy levels and the diversity of
transitions in the transition maps. A highly organized team (lower right), as
typified by a SPAN team under stress, is shown by tightly-organized transitions
and low entropy levels, equivalent to the random usage of only nine of the 25
NS_E symbols. NS transitions pooled from the Scenario segments of six SOAC
teams still show restricted transitions, but the mean entropy has increased. As
teams progress after their initial training and develop more experience (SUB
Teams), the entropy levels and the diversity of the transitions further increase;
from the performance metric, this stage would approximate the ‘sweet spot’ of
team function. The data from zero-history student teams who had not worked
together (lower left), and were unfamiliar with both the task and domain,
showed the highest entropy. Their entropy levels were nearly equivalent to
randomized NS_E data streams. As discussed in the Introduction, this
hypothesized structure is consistent with the idea that teams, like many complex
systems, are thought to operate at an organization level between random and
highly
-
From
a
compl
e
p
ast t
h
SPAN
nearly
SUB t
e
Fig. 11
exhibi
t
norma
l
hetero
g
shows
and a
d
and o
r
naviga
t
the tea
m
regula
t
organi
z
fractal
spots (
M
avenu
e
trainin
g
the fu
n
p
erfor
m
NDPLS,
1
-
organized, at
a
complexity
p
e
xity (Crutchf
i
h
at is needed
t
teams under
s
random proc
e
e
ams would ha
v
. A model of e
x
This divers
i
t
a modest ran
g
l
ly functioning
g
eneous in its
increased coo
r
d
ecrease in fi
n
r
der in the s
y
t
ion teams ma
y
m
demonstrat
e
t
ion. As team
n
z
ed around ep
scaling analy
s
M
uzy, Bacry,
&
We propos
e
e
for the deve
l
g
activities in
n
ctional statu
s
m
ance or de
c
1
7(1), Or
g
ani
z
the so-called
p
erspective, Fi
g
i
eld & Young
,
t
o predict the
s
tress would
h
e
ss and the o
t
v
e the highest
x
pertise and t
h
i
ty of organi
z
g
e of comple
x
market is cha
o
sources and
f
r
dinated beha
v
n
ancial diversi
t
y
stem process
y
function clo
s
e
s both stabilit
y
n
eurosynchron
i
isodes or per
t
s
is may provi
d
&
Arneodo, 1
9
e
that the stu
d
l
opment of ad
complex env
i
s
of a team
i
c
isions and t
o
z
ational Neur
o
edge of cha
o
g
. 11 can be t
h
,
1989), whic
h
future. In pa
r
h
ave low stati
s
t
her highly o
r
complexity.
h
e cognitive or
g
z
ational state
s
x
states, not u
n
o
tic, and the l
o
f
lows of infor
m
v
ior of a large
n
t
y (Sornette, 1
leads to a “
c
s
er to the “sw
e
y
and flexibili
t
i
es also fluctu
a
t
urbations wit
h
d
e an approach
9
93).
d
ies that we h
a
aptive trainin
g
i
ronments is t
h
i
n order to a
s
o
adaptively
o
d
y
namics o
f
T
o
s or self-org
a
h
ought of in t
e
h
is the infor
m
r
ticular, both
s
tical comple
x
r
ganized, whe
r
g
anization of S
s
suggests th
a
n
like stock ma
r
o
cal variabilit
y
m
ation. But a
n
umber of age
n
998). This ad
d
c
rash.” Simil
a
e
et spot” of or
g
t
y in the form
o
a
te on much s
m
h
in a single t
a
h
for better de
f
a
ve presented
g
systems. A
c
h
e ability to r
a
s
sess the qua
l
rearrange th
e
T
eams
a
nized critical
i
e
rms of statisti
c
m
ation about
t
zero-history
a
x
ity, one bein
g
r
eas experien
c
PAN teams.
a
t SPAN tea
m
r
ket volatility.
y
of the proces
s
market in cr
i
nts in the mar
k
d
itional struct
u
a
rly, experien
c
g
anization wh
e
o
f supportive
c
m
aller timescal
a
sk performan
c
f
ining such sw
e
here suggest
c
ommon goal
a
pidly determ
i
l
ity of a tea
m
e
team or t
a
83
i
ty.
c
al
t
he
a
nd
g
a
c
ed
m
s
A
s
is
i
sis
k
et
u
re
c
ed
e
re
c
o-
es,
c
e,
e
et
an
of
i
ne
m
s’
a
sk
84 NDPLS, 17(1), Stevens et al.
components to better optimize the team. One of the challenges in accomplishing
this goal is the development of rapid, relevant, and reliable models for providing
this information to the trainers and trainees. With the creation of standardized
models of NS_E expression (Stevens, Galloway, Wang, Berka, & Behneman,
2011) it may now be possible to direct real-time EEG streams into our modeling
system and rapidly report back the entropy and attractor basin status of the team.
ACKNOWLEDGMENTS
This work was supported in part by The Defense Advanced Research
Projects Agency under contract number(s) W31P4Q12C0166, and NSF SBIR
grants IIP 0822020 and IIP 1215327. The views, opinions, aor findings
contained are those of the authors and should not be interpreted as representing
the official views or policies, either expressed or implied, of the Defense
Advanced Research Projects Agency or the Department of Defense. Special
thanks to Chris Berka, Adrienne Behneman, and Veasna Tan for EEG technical
support and assistance with the data collection.
REFERENCES
Ahire, S. L. (1997). Management science – Total quality management interfaces: An
integrative framework. Interfaces, 27, 91-105.
Ashenfelter, K. (2007) Simultaneous analysis of verbal and nonverbal data during
conversation: symmetry and turn-taking. Unpublished thesis, University of
Notre Dame.
Bak, P., Tang, C., & Wiesenfeld, K. (1987). Self-organized criticality: an explanation of
1/f noise. Physical Review Letters, 59, 381–384.
Berka, C., Levendowski, D. J., Cvetinovic, M. M., Petrovic, M. M., Davis G., &
Lumicao, M.N. (2004). Real-time analysis of EEG indexes of alertness,
cognition, and memory acquired with a wireless EEG headset. International
Journal of Human-Computer Interaction, 17, 151-170.
Crutchfield, J. P. & Young, K. (1989). Inferring statistical complexity. Physical Review
Letters, 63, 105-108.
Daw, C. S., Finney, C. E. A., & Tracy, E. R. (2003). A review of symbolic analysis of
experimental data. Review of Scientific Instruments. 74, 915-933.
Drew, P. (2005). Conversation analysis. In Fitch & Sanders (Eds.), Handbook of
language and social interaction (pp. 71-102). Mahwah, NJ: Lawrence
Erlbaum.
Dumas, G., Nadel, J., Soussignan, R., Martinerie, J., & Garnero, L. (2010). Interbrain
synchronization during social interaction. PlosOne, 5(8) e12166. DOI: 10
1371/journal.pone0012166.
Gorman, J. C., Amazeen, P. G., & Cooke, N. J (2010). Team coordination dynamics.
Nonlinear Dynamics, Psychology, and Life Sciences, 14, 265-289.
Guastello, S.J., Pincus, D., & Gunderson, P.R. (2006). Electrodermal arousal between
participants in a conversation: Nonlinear dynamics for linkage effects.
Nonlinear Dynamics, Psychology and Life Sciences, 10, 365-399.
NDPLS, 17(1), Organizational Neurodynamics of Teams 85
Howard, B. (1996, February). Cognitive engagement in cooperative learning. Annual
Meeting of the Eastern Educational Research Association. Boston, MA.
Retrieved June 8, 2011 from
http://www.eric.ed.gov/ERICWebPortal/search/detailmini.jsp?_nfpb=true&_&ERICExtS
earch_SearchValue_0=ED404352&ERICExtSearch_SearchType_0=no&accno=ED4043
52. Accessed 8 June 2011.
Huang, R. S., Jung, T. P., & Makeig, S. (2007). Multi-scale EEG brain dynamics during
sustained attention tasks. Proceedings of the 2007 IEEE International
Conference on Acoustics, Speech and Signal Processing (ICASSP2007) (pp.
1173-1176). Honolulu, HI: IEEE.
Jung, T. P., Makeig, S., Stensmo, M., & Sejnowski, T. J. (1997). Estimating alertness
from the EEG power spectrum. IEEE Transactions on Biomedical Engineering,
44, 60-69.
Kelly, S. P. Docktree, P., Reilly, R. B., & Robertson, I. H. (2003). EEG Alpha power and
coherence time courses in a sustained attention task. Proceedings of the
International Conference on Neural Engineering, March 20-22 (pp. 83-86).
Red Hook, NY: Curran Associated, Inc.
Kozlowski, J S. & Ilgen, D. R. (2006). Enhancing the effectiveness of work groups and
teams. Psychological Science in the Public Interest, 7, 77-124.
Muzy, J. F., Bacry, E., & Arneodo, A. (1993). Multifractal formalism for fractal signals:
The structure-function approach versus the wavelet-transform modulus-maxima
method. Physical Review E, 47, 875-884.
Pincus, D., (2009). Coherence, complexity and information flow In S. Guastello, M.
Koopmans, & D. Pincus (Eds.), Chaos and Complexity in Psychology (pp. 335-
369). New York: Cambridge University Press.
Roberts, E. & Young, R.M (2008). Maintaining cognitive engagement in training
scenarios using explicit cognitive models. Interservice/Industry Service
Training and Education Conference (I/ITSEC) (paper 8217). National Training
and Simulation Association, Arlington, VA.
Salem, P. (2011). The sweet spots in human communication. Nonlinear Dynamics,
Psychology and Life Sciences, 15, 389-406.
Shockley, K., Santana, M.-V., & Fowler, C. A. (2003). Mutual interpersonal postural
constraints are involved in cooperative conversation. Journal of Experimental
Psychology: Human Perception and Performance, 29, 326–332.
Shannon, C. & Weaver, W. (1949). The mathematical theory of communication. Urbana,
IL: University of Illinois Press.
Sornette, D. (1998). Critical market crashes. Physics Reports, 78, 1-98.
Stephens, G., Silbert, L., & Hasson, U. (2010). Speaker-listener neural coupling underlies
successful communication. Proceedings of the. National Academy of. Sciences,
106, 10841-10846.
Stevens, R. H., Galloway, T., Berka, C., & Sprang, M. (2009). Can neurophysiologic
synchronies be detected during collaborative teamwork? In D. Schmorrow, I.
Estabrook, & M. Grootjen (Eds.), Frontiers of Augmented Cognition:
Proceedings: HCI International 2009 (pp. 271-275). Berlin: Springer-Verlag.
Stevens, R., Galloway, T., Berka, C., Behneman, A., Wohlgemuth, T., et al. (2011).
Linking models of team neurophysiologic synchronies for engagement and
workload with measures of team communication. In Proceedings of the 20th
Conference on Behavior Representation in Modeling and Simulation. (11-
BRIMS-019). Centerville, OH: The BRIMS Society.
86 NDPLS, 17(1), Stevens et al.
Stevens, R., Galloway, T., Wang, P., Berka, C., & Behneman, A. (2011). Developing
systems for the rapid modeling of team neurodynamics. In Interservice /
Industry Training Simulation and Education Conference (I/ITSEC) 2011 (Paper
No. 11097). National Training and Simulation Association, Arlington, VA.
Stevens, R. & Gorman, G. (2011). Mapping cognitive attractors onto the dynamic
landscapes of teamwork. In D. Schmorrow & C. Fidopiastis (Eds.),
Proceedings of the 14th International Conference on Human-Computer
Interaction (pp. 366-375). Berlin: Springer-Verlag.
Stevens, R.H., Galloway, T., Wang, P., & Berka, C. (2012). Cognitive neurophysiologic
synchronies: What can they contribute to the study of teamwork? Human
Factors, 54, 489-502.
Treffner, P. J. & Kelso, J. A. S. (1999). Dynamic encounters: Long-memory during
functional stabilization. Ecological Psychology, 11, 103-137.
Warner, N., Letsky, M., & Cowen, M. (2005). Cognitive model of team collaboration:
macro-cognitive focus. In Proceedings of the 49th Human Factors and
Ergonomics Society Annual Meeting (pp. 269-273). Santa Monica, CA: Human
Factors and Ergonomics Society.
Webber, C. L. & Zbilut, J. P. (1994). Dynamical assessment of physiological systems
and states using recurrence plot strategies. Journal of Applied Physiology, 76,
965-973.
... The neurodynamics approach is a more recent perspective on team cognition and has been explored primarily by Stevens & colleagues (e.g. Stevens et al., 2010, 2012Stevens, 2013;Stevens, Galloway and Lamb, 2014;Stevens and Galloway, 2016;Willemsen-Dunlap, 2017, 2018;Stevens, Galloway and Willemson-Dunlap, 2017;Stevens, Willemsen-Dunlap, et al., 2018;Stevens, Galloway, et al., 2018). This perspective views teams as complex systems which operate at a level which self-organises between random and highly organised states (Stevens, 2013). ...
... Stevens et al., 2010, 2012Stevens, 2013;Stevens, Galloway and Lamb, 2014;Stevens and Galloway, 2016;Willemsen-Dunlap, 2017, 2018;Stevens, Galloway and Willemson-Dunlap, 2017;Stevens, Willemsen-Dunlap, et al., 2018;Stevens, Galloway, et al., 2018). This perspective views teams as complex systems which operate at a level which self-organises between random and highly organised states (Stevens, 2013). Teamwork in this sense was first described as "the continuous effort involved in stabilization of an inherently unstable system" (Gorman et al., 2010;Stevens, 2013;Treffner and Kelso, 1999) and more recently the "evolving dynamics across temporal scales that are continually punctuated by small and large fluctuations as disturbances to the team's normal rhythms are encountered and resolved" . ...
... This perspective views teams as complex systems which operate at a level which self-organises between random and highly organised states (Stevens, 2013). Teamwork in this sense was first described as "the continuous effort involved in stabilization of an inherently unstable system" (Gorman et al., 2010;Stevens, 2013;Treffner and Kelso, 1999) and more recently the "evolving dynamics across temporal scales that are continually punctuated by small and large fluctuations as disturbances to the team's normal rhythms are encountered and resolved" . In this sense, the dynamics of teams not only includes the communication and behaviours of teams seen in the THEDA perspective, it also considers neurophysiological changes. ...
Thesis
Full-text available
For decades both industry and academia have been interested in situation awareness, from individual situation awareness to system situation awareness of dynamic collaborative systems. Several theories and definitions exist for situation awareness and although considerable research has been conducted in this domain no definitive consensus has been reached. Therefore, the purpose of the research in this thesis is not develop new theories or definitions, but to explore how situation awareness presents itself in teams and systems in terms of team cognition. The methods used in this thesis include simulating team tasks using agent-based modelling, analysing team knowledge using concept maps and analysing team processes using entropy. In order to remove the risk of intrusion on the tasks being explored, the communications of team members are recorded and used as the primary data for the analyses conducted. Visually presenting knowledge of agents using concept maps made it easier to understand how the information was stored and transferred throughout the teams. An interesting result showed that it was not important for all agents to have the same information when key decisions were made and that when information is not shared the team performed better and with greater accuracy than when there was a focus on information sharing. Visually presenting team processes using entropy and process distribution allowed for patterns of behaviour to be identified. Results show that while individuals within teams feel confident with the amount of knowledge they have they will focus on working independent up until the point they can no longer achieve results on their own, at that point the team shifts to teamworking. The differences between teamwork and taskwork are related to the theories of shared and distributed situation awareness, concluding that shifts in team processes represent shifts in the two types of situation awareness.
... Since it has been observed, that physiological changes can synchronise amongst small group members and have been found in relation to group performances (see, e.g. Stevens et al., 2012;Stevens, Amazeen, and Likens, 2013;Berka and Stikic, 2017), it would be interesting to analyse, if, for instance, HiBeta synchronisation is related to increases in flow synchronisation (i.e. correlation over time) or clustering (i.e. ...
... Further evaluation of this relationship will likely have to venture into the domain of neurophysiological synchronisation research (see, e.g. Palumbo et al., 2017;Stevens, Amazeen, and Likens, 2013;Berka and Stikic, 2017). In synchronisation research, is has been observed that neurophysiological signals of multiple individuals occasionally show correlations over time and that such synchronisations can relate to superior group performances and social processes like emotional contagion (Labonté-Lemoyne et al., 2016) or stress-buffering (Tse et al., 2016). ...
... Such synchronisations have previously been reported in relation to performances of or leadership emergence in small groups (see, e.g. Stevens et al., 2012;Stevens, Amazeen, and Likens, 2013;Berka and Stikic, 2017). As KW is increasingly conducted in small groups, the integration of such multi-person data into a flow-detection system seems promising. ...
Thesis
Full-text available
The experience of flow is a unique sensation of complete task absorption and effortless action that is highlighted as a correlate of peak performances, personal and social growth, and general well-being. For organisations, higher flow frequencies, therefore, relate to a more engaged, skilled, and productive workforce. Especially as global phenomena like increasing knowledge work demand and low worker engagement are developing, organisations could strongly benefit from fostering workers’ flow experiences. However, facilitating flow represents a substantial challenge due to the variety of workers’ abilities, tasks and workplace configurations. Knowledge workers are faced with unstructured and complex tasks, that require numerous domain-specific abilities and cooperation with others. Workplaces are diversifying with boundaries disappearing between centralized and digitally-mediated workspaces. This variety means that only person-, task- and situation-independent approaches can deliver comprehensive flow support. For this reason, research on the experiences neurophysiological basis is increasingly pursued. On this basis, adaptive Neuro-Information Systems (NeuroIS) could be developed that are able to detect flow continuously (especially through wearable sensor systems), and that can provide flow-supporting mechanisms. Presently, despite these efforts, the knowledge on how to detect flow with neurophysiological measures is sparse, highly fragmented, and lacks experimental variety. On the individual level, competing propositions exist that have not been consolidated through cross-situational, and multi-sensor observation. On the group level, almost no research has been conducted to investigate neurophysiological correlates in social interactions, particularly not in digitally-mediated interactions. This dissertation addresses these gaps through the cross-situational observation of flow using wearable ECG and EEG sensor systems. In doing so, limitations in the present state of experimental flow research are addressed that refer to central shortcomings of established paradigms for the controlled elicitation of flow experiences. Specifically, two experiments are conducted with manipulations of difficulty, naturalism, autonomy, and social interaction to investigate the question of how flow elicitation can be intensified, and the experience detected more robustly across situations. These investigations are based on an extensive integration of the theoretic and empiric literature on flow neurophysiology. Altogether, the results suggest flow to be represented by moderate physiological activation and mental workload, by increased attentional task engagement and by affective neutrality. Especially EEG features indicate a diagnostic potential to separate lower from higher flow intensities by the reflection of optimal and non-optimal (individual and group) task difficulties. To catalyse, that the positive promises of fostering flow in individuals and social units, can be realised, avenues to advance flow facilitation research are outlined.
... Better work performance outcomes would also be expected when teams are similarly synchronized (Elkins et al., 2009;Stevens et al., 2013). ...
Preprint
Full-text available
We introduce "entanglement", a novel metric to measure how synchronized communication between team members is. This measure calculates the Euclidean distance among team members' social network metrics timeseries. We validate the metric with four case studies. The first case study uses entanglement of 11 medical innovation teams to predict team performance and learning behavior. The second case looks at the e-mail communication of 113 senior executives of an international services firm, predicting employee turnover through lack of entanglement of an employee. The third case analyzes the individual employee performance of 81 managers. The fourth case study predicts performance of 13 customer-dedicated teams at a big international company by comparing entanglement in the e-mail interactions with satisfaction of their customers measured through Net Promoter Score (NPS). While we can only speculate about what is causing the entanglement effect, we find that it is a new and versatile indicator for the analysis of employees' communication, analyzing the hitherto underused temporal dimension of online social networks which could be used as a powerful predictor of employee and team performance, employee turnover, and customer satisfaction.
... While these studies only look at synchronization as neuromuscular coordination and task coordination, there are research efforts currently underway to uncover connections between synchronization in cognition, task structures, and performance outcomes in teams (Gipson et al., 2016). Better work performance outcomes would also be expected when teams are similarly synchronized (Elkins et al., 2009;Stevens et al., 2013). ...
Article
We introduce “entanglement”, a novel metric to measure how synchronized communication between team members is. This measure calculates the Euclidean distance among team members’ social network metrics timeseries. We validate the metric with four case studies. The first case study uses entanglement of 11 medical innovation teams to predict team performance and learning behavior. The second case looks at the e-mail communication of 113 senior executives of an international services firm, predicting employee turnover through lack of entanglement of an employee. The third case analyzes the individual employee performance of 81 managers. The fourth case study predicts performance of 13 customer-dedicated teams at a big international company by comparing entanglement in the e-mail interactions with satisfaction of their customers measured through Net Promoter Score (NPS). While we can only speculate about what is causing the entanglement effect, we find that it is a new and versatile indicator for the analysis of employees’ communication, analyzing the hitherto underused temporal dimension of online social networks which could be used as a powerful predictor of employee and team performance, employee turnover, and customer satisfaction.
... It has been proposed that biological systems, like teams, are hierarchies of information that are functionally organized across spatial and time scales (Flack, 2017a). Uncertainty is the messenger on this hierarchy guiding information back and forth between the environment and the team (Flack et al., 2012), with ripples and islands in these information streams representing periods of changing organization [see Stevens et al. (2013) for team examples]. This changing information helps the brain identify regularities in the environment and use them to shape adaptations along the macroscopic and microscopic continuum of experience and learning (Daniels et al., 2017). ...
Article
Full-text available
Neurodynamic organizations are information-based abstractions, expressed in bits, of the structure of long duration EEG amplitude levels. Neurodynamic information (NI, the variable of neurodynamic organization) is thought to continually accumulate as EEG amplitudes cycle through periods of persistent activation and deactivation in response to the activities and uncertainties of teamwork. Here we show that (1) Neurodynamic information levels were a better predictor of uncertainty and novice and expert behaviors than were the EEG power levels from which NI was derived. (2) Spatial and temporal parsing of team NI from experienced submarine navigation and healthcare teams showed that it was composed of discrete peaks with durations up to 20–60 s, and identified the involvement of activated delta waves when precise motor control was needed. (3) The relationship between NI and EEG power was complex varying by brain regions, EEG frequencies, and global vs. local brain interactions. The presence of an organizational system of information that parallels the amplitude of EEG rhythms is important as it provides a greatly reduced data dimension while retaining the essential system features, i.e., linkages to higher scale behaviors that span temporal and spatial scales of teamwork. In this way the combinatorial explosion of EEG rhythmic variables at micro levels become compressed into an intermediate system of information and organization which links to macro-scale team and team member behaviors. These studies provide an avenue for understanding how complex organizations arise from the dynamics of underlying micro-scale variables. The study also has practical implications for how micro-scale variables might be better represented, both conceptually and in terms of parsimony, for training machines to recognize human behaviors that span scales of teams.
... Now it is possible to understand change of leadership as well as organization learning based on registration of micro interactions [53]. More complex simulation experiments with telemetric EEG allow associating brain activity features with output of teamwork [53] [54]. On the other hand, measurement of mechanical quasi-oscillations is useful for rehabilitation research, since it allows evaluating the overall change in a system's functional state [55]. ...
Article
Full-text available
The main features of early rehabilitation after severe brain damage are discussed in the article. The most important component for the entire rehabilitation process and the subsequent life of the patient is considered restoration of consciousness. Team seems to be a key factor in regaining consciousness along with the restoration of vital functions, movement, cognition, and behavior in these patients. The basic working principle is feedback to any minimal movement, or vegetative signal of a patient, beyond specific professional targets. A network of feedbacks with a patient and between professionals , that is, free flow of information, can be built only through work in a transdisciplinary team mode. The net of feedbacks with the patient and inter professional ones builds up the team as Non-linear Complex System. Characteristics of "Team-Patient" system status are energy, entropy, and complexity. Teamwork techniques are individualized for resulting optimization of system condition. Increase of complexity is a powerful tool for propulsion of recovery process. Then consciousness may appear as a result of system self-organization. The article reflects the authors' view on interdiscip-linary studies of the phenomenon of consciousness through its impairment and recovery. It focuses on the work of the "proper rehabilitation team", the mechanisms of its action and methods for researching the occurring phenomena .
... Now it is possible to understand change of leadership as well as organization learning based on registration of micro interactions [53]. More complex simulation experiments with telemetric EEG allow associating brain activity features with output of teamwork [53] [54]. On the other hand, measurement of mechanical quasi-oscillations is useful for rehabilitation research, since it allows evaluating the overall change in a system's functional state [55]. ...
Article
Full-text available
The main features of early rehabilitation after severe brain damage are discussed in the article. The most important component for the entire rehabilitation process and the subsequent life of the patient is considered restoration of consciousness. Team seems to be a key factor in regaining consciousness along with the restoration of vital functions, movement, cognition, and behavior in these patients. The basic working principle is feedback to any minimal movement, or vegetative signal of a patient, beyond specific professional targets. A network of feedbacks with a patient and between professionals, that is, free flow of information, can be built only through work in a transdisciplinary team mode. The net of feedbacks with the patient and inter-professional ones builds up the team as Non-linear Complex System. Characteristics of “Team-Patient” system status are energy, entropy, and complexity. Teamwork techniques are individualized for resulting optimization of system condition. Increase of complexity is a powerful tool for propulsion of recovery process. Then consciousness may appear as a result of system self-organization. The article reflects the authors’ view on interdisciplinary studies of the phenomenon of consciousness through its impairment and recovery. It focuses on the work of the “proper rehabilitation team”, the mechanisms of its action and methods for researching the occurring phenomena.
... Now it is possible to understand change of leadership as well as organization learning based on registration of micro interactions [53]. More complex simulation experiments with telemetric EEG allow associating brain activity features with output of teamwork [53] [54]. On the other hand, measurement of mechanical quasi-oscillations is useful for rehabilitation research, since it allows evaluating the overall change in a system's functional state [55]. ...
Article
Full-text available
Objectives. To study the characteristics of impairments to brain activity in patients after craniocerebral trauma (CCT) on performing double tasks as compared with normal subjects. Materials and methods. Clinical electroencephalographic (EEG) and stabilographic investigations were run in nine patients (mean age 25.0 ± 1.2 years) within three months of moderate CCT. The control group consisted of 18 healthy subjects (mean age 26.6 ± 0.07 years). All subjects performed two motor and two cognitive tasks separately and simultaneously, i.e., dual tasks. Results. Patients with CCT had cognitive deficit with retention of posture control. EEG data showed that performance of the cognitive components of dual tasks was accompanied by decreases in the coherence of the slow rhythms in the left hemisphere and frontal areas. Healthy subjects showed increases in coherence in the slow EEG rhythms in these parts of the brain. Conclusions. Dual tasks provided an informative method for evaluating the predominance of cognitive deficit after moderate CCT and use of this approach as a rehabilitation measure promoted clinical improvements.
Article
Uncertainty is a fundamental property of neural computation that becomes amplified when sensory information does not match a person’s expectations of the world. Uncertainty and hesitation are often early indicators of potential disruption, and the ability to rapidly measure uncertainty would have implications for future educational and training efforts by targeting reflective discussions about past actions, supporting in-progress corrections, and generating forecasts about future disruptions. An approach is described combining neurodynamics and machine learning to provide quantitative measures of uncertainty. Models of neurodynamic information derived from electroencephalogram (EEG) brainwaves have provided detailed neurodynamic histories of US Navy submarine navigation team members. Persistent periods (25–30 s) of neurodynamic information were seen as discrete peaks when establishing the submarine’s position and were identified as periods of uncertainty by an artificial intelligence (AI) system previously trained to recognize the frequency, magnitude, and duration of different patterns of uncertainty in healthcare and student teams. Transition matrices of neural network states closely predicted the future uncertainty of the navigation team during the three minutes prior to a grounding event. These studies suggest that the dynamics of uncertainty may have common characteristics across teams and tasks and that forecasts of their short-term evolution can be estimated.
Article
Full-text available
The True Nature of Psychotherapy What is psychotherapy? This simple question can stimulate hours of discussion, producing more new questions for each new answer. For example, one may ask if there is a single answer to what psychotherapy is, or if in fact psychotherapy is always different for each unique client. Probing deeper, one may also wonder if scientifically informed “theories” of psychotherapy really add to the “truth” of what therapy is for any given client, or whether truth is actually derived through a constructive process. Such questions raise the question of the merits of “truth” versus “utility,” as even the most scientifically controlled or “manualized” approaches to treatment encourage the therapist to accommodate any approach to fit the values, goals, and unique situations of the client. It appears that defining psychotherapy is a difficult matter, creating practical and scientific questions that are worthy of deep exploration. Psychotherapy has a great tradition of such deep self-examination. For example, Frank and Frank (1991) have influenced two generations of therapists with the well-reasoned yet apparently controversial suggestion that modern psychotherapy is based more on applied rhetoric than on scientific discovery. The psychotherapy field has generally been forced to conclude that factors common across approaches are better predictors of treatment outcome than are the use of a specific technique (Orlinsky & Howard, 1995), particularly the quality of the therapist-client relationship (Orlinsky, Grawe, & Parks, 1994).
Article
Full-text available
The purpose of this paper is to describe a cognitive model of team collaboration emphasizing the human decision-making processes used during team collaboration. The descriptive model includes the domain characteristics, collaboration stages, meta-and macro cognitive processes and the mechanisms for achieving the stages and cognitive processes. Two experiments were designed to provide empirical data on the validity of the collaboration stages and cognitive processes of the model. Both face-to-face and asynchronous, distributed teams demonstrated behavior that supports the existence of the collaboration stages along with seven cognitive processes.
Article
Full-text available
We explored the problem of actively stabilizing an inherently unstable system. A method for analyzing scale invariant time series was used to examine hand displacement and rod angle of an inverted pendulum actively balanced by normal individuals. Estimation of the Hurst exponent revealed long term fractal correlations for more than 3 min into the past. Two regimes, a short-term time scale of positive correlation (persistence) and a long term time scale of negative correlation (antipersistence), were revealed, and their transition points were modulated by the rod's natural frequency. Such nonlocal temporal structure constitutes a physical characteristic known as long memory and may be considered inherent to the coordination dynamics itself. The implications of long-memory processes are discussed in terms of the dynamical mechanisms underlying ecological encounters and how affordances require a persisting environment for their definition.
Teams of people working together for a common purpose have been a centerpiece of human social organization ever since our ancient ancestors first banded together to hunt game, raise families, and defend their communities. Human history is largely a story of people working together in groups to explore, achieve, and conquer. Yet, the modern concept of work in large organizations that developed in the late 19th and early 20th centuries is largely a tale of work as a collection of individual jobs. A variety of global forces unfolding over the last two decades, however, has pushed organizations worldwide to restructure work around teams, to enable more rapid, flexible, and adaptive responses to the unexpected. This shift in the structure of work has made team effectiveness a salient organizational concern. Teams touch our lives everyday and their effectiveness is important to well-being across a wide range of societal functions. There is over 50 years of psychological research—literally thousands of studies—focused on understanding and influencing the processes that underlie team effectiveness. Our goal in this monograph is to sift through this voluminous literature to identify what we know, what we think we know, and what we need to know to improve the effectiveness of work groups and teams. We begin by defining team effectiveness and establishing the conceptual underpinnings of our approach to understanding it. We then turn to our review, which concentrates primarily on topics that have well-developed theoretical and empirical foundations, to ensure that our conclusions and recommendations are on firm footing. Our review begins by focusing on cognitive, motivational/affective, and behavioral team processes—processes that enable team members to combine their resources to resolve task demands and, in so doing, be effective. We then turn our attention to identifying interventions, or “levers,” that can shape or align team processes and thereby provide tools and applications that can improve team effectiveness. Topic-specific conclusions and recommendations are given throughout the review. There is a solid foundation for concluding that there is an emerging science of team effectiveness and that findings from this research foundation provide several means to improve team effectiveness. In the concluding section, we summarize our primary findings to highlight specific research, application, and policy recommendations for enhancing the effectiveness of work groups and teams.
Article
Management science (MS) and total quality management (TQM) are powerful approaches to improving the operations of organizations. While MS uses quantitative techniques to yield information-based strategies for decision problems, TQM aims at improving the quality of processes to produce superior products at lower cost. Hence, MS and TQM should, in principle, be intertwined to attain efficient and cost-effective operations. However, neither researchers nor practitioners in these two fields have recognized the vital linkage. By viewing TQM from process-improvement and implementation perspectives, I establish links of prominent MS techniques, such as decision analysis, forecasting, mathematical programming, and simulation, to various facets of TQM. Applications of two important tools (linear programming and simulation) to the service operations of high customer-contact organizations illustrate these linkages. I strongly encourage organizations to systematically integrate MS techniques into TQM programs. Such efforts will result in better planning and implementation of TQM efforts.