ArticlePDF Available

Measuring Team Creativity Through Longitudinal Social Signals

Authors:

Abstract and Figures

Research into human dynamical systems has long sought to identify robust signals for human behavior. We have discovered a series of social network-based indicators that are reliable predictors of team creativity and collaborative innovation. We extract these signals from electronic records of interpersonal interactions, including e-mail, and face-to-face interaction measured via sociometric badges. The first of these signals is Rotating Leadership, measuring the degree to which, over time, actors in a team vary in how central they are to team's communication network's structure. The second is Rotating Contribution, which measures the degree to which, over time, actors in a team vary in the ratio of communications they distribute versus receive. The third is Prompt Response Time, which measures, over time, the responsiveness of actors to one another's communications. Finally, we demonstrate the predictive utility of these signals in a variety of contexts, showing them to be robust to various methods of evaluating innovation.
Content may be subject to copyright.
Measuring Team Creativity Through Longitudinal Social Signals
Peter A. Gloor, Adam Almozlino, Orr Inbar
MIT Center for Collective Intelligence
5 Cambridge Center, Cambridge MA 02139, USA
pgloor@mit.edu
Wei Lo
Computer Science and Technology department
Zhejiang University, Hangzhou, P.R. China
Shannon Provost
McCombs School of Business, University of Texas at Austin
Austin, TX, USA
Summary
Research into human dynamical systems has long sought to identify robust signals for human
behavior. We have discovered a series of social network-based indicators that are reliable
predictors of team creativity and collaborative innovation. We extract these signals from
electronic records of interpersonal interactions, including e-mail, and face-to-face interaction
measured via sociometric badges. The first of these signals is Rotating Leadership, measuring the
degree to which, over time, actors in a team vary in how central they are to team’s communication
network’s structure. The second is Rotating Contribution, which measures the degree to which,
over time, actors in a team vary in the ratio of communications they distribute versus receive. The
third is Prompt Response Time, which measures, over time, the responsiveness of actors to one
another’s communications. Finally, we demonstrate the predictive utility of these signals in a
variety of contexts, showing them to be robust to various methods of evaluating innovation.
Introduction
In this paper we introduce a series of longitudinal, network-based measures of social interaction patterns
that predict collaborative innovation. Innovation is a universal, emergent human behavior. According to
noted evolutionary biologist E.O Wilson “…it was necessary for the evolving populations to acquire an
ever higher degree of intelligence. They had to feel empathy for others, to measure the emotions of
friends and enemy alike, to judge the intentions of all of them, and to plan a strategy for personal social
interactions” [1]. Innovation is a universal, emergent human behavior, one that rarely occurs through the
actions of a single individual, but rather through collaboration among individuals [2]. Here we focus on
the predictive utility of observing this collaboration at the level of interpersonal interaction events.
Recently, researchers have had success in identifying reliable, quantitative indicators of phenomena in
human systems. Among these indicators are “honest signals” [3][4][5], which signify the presence of
social influence. This name captures both the separation of these signals' from the subjectivity that often
plagues other methods for measuring human behavior, and the robustness of these signals to a variety of
behavioral contexts. Understanding these “honest signals” can convey a significant advantage. To quote
E.O Wilson again, “…social intelligence was therefore always at a high premium. A sharp sense of
empathy can make a huge difference and with it an ability to manipulate, to gain cooperation, and to
deceive[1]. Robust, quantitative measures for collective human behavior may serve as the quantitative,
larger-scale analog for individual social intelligence.
Previous work studying collective creativity and innovation has been primarily qualitative, focused on the
creativity of individuals, or both [6][7]. Other research has been restricted to a particular interpretation of
creativity, studying for example patent production [8], or to a particular setting, studying for example
large corporations [9]. Therefore, this research has failed to identify reliable signals of collective
innovation.
Part of the reason previous work has had limited success may lie in the difficulty of understanding
innovation itself. A formal definition of innovation remains elusive, as does the boundary between
incremental improvements and innovative change. If a certain dependent variable, such as creativity, is
difficult to formally define, it may be difficult to identify a quantitative and reproducible independent
variable that indicates the dependent.
Our Approach
We have attempted to work around this issue by evaluating several different proxies for creativity across
several different scenarios, and identifying measures that reliably signal the presence of these proxies
across the scenarios. Using a wide selection of proxies in a variety of context, we have identified
reproducible independent variables that strongly correlate with the proxies. We term these variables (1)
Rotating Leadership, (2) Rotating Contribution, (3) Prompt Response Time.
From these variables, Rotating Leadership and Rotating Contribution show positive correlation in
“creative” work scenarios, but strong negative correlation with “non-creative” scenarios, suggesting that
Rotating Leadership and Rotating Contribution are a good “honest signal” for team creativity. This
corresponds with the intuition that creative work requires innovation and breaking known patterns of
thought and behavior, while breaking known patterns may disrupt non-creative work. Prompt Response
Time, on the other hand, shows positive correlation across all scenarios, suggesting that it is a better
indicator of team productivity. This corresponds with the intuition that it is, in general, better to have a
more promptly communicating team.
(1) Rotating Leadership (RL)
Rotating Leadership (RL) measures the degree to which, over time, the members in a team vary in how
“central” they are to the team’s communications. The advantage of centralized leadership for creative
tasks was for instance observed among Wikipedians [10], where it was found that Wikipedia articles
authored by more centrally communicating editors became articles of the highest quality (featured
articles) more rapidly. RL can be observed in a visualization of a network when distinct nodes appear,
over time, to oscillate between central and peripheral positions in the network. Intuitively, RL evaluates
how much, across time and the team members, team members switch between being highly central to the
overall communications of the team, and being peripheral to those communications. Formally, RL
measures oscillations in Betweenness Centrality (BC) over time among actors in the team.
The effects of the centrality of team’s actors to the team’s performance was first observed among teams
of Eclipse open source developers communicating electronically [12]. It was subsequently observed in a
study of a marketing team in a bank communicating face-to-face [13], and in a study of nurses
communicating in a hospital [14]. In this last scenario, quantitative measures were compared with
personality characteristics such as openness, as measured by the Neo-FFI [15], and group creativity was
measured through peer and management/instructor assessment, based on the premise that experts can
identify creativity [7]. Note that teams composed of highly intelligent individuals are not necessarily
intelligent as a whole [16], while measures such as RL were dependably correlated with team creativity.
Betweenness centrality [11] (BC) is a global measure of how centrally located a node is in the structure of
a network. For a given node, it is measured by evaluating the shortest paths in the network, specifically,
the proportion of all shortest paths in the network that pass through the node of interest. Mathematically,
BC of a node v is defined as:
where is the total number of shortest paths from node to node and is the number of those
paths that pass through v.
In order to calculate RL, it is necessary to aggregate measures of BC, which occur at the scale of an
individual actor at an individual time step, to the scale of the whole network over the full time frame. In
order to do this in a fashion that indicates variation in BC we counted the number of local maxima and
minima in the vector of each actor’s BC over time, and then summed this number across the actors in a
team.
Formally, we count the local maxima of function f(t)=g(t) within time interval [t1,t2]. There is a local
maximum for time t at point t*, if there exists some ε > 0 such that f(t*) f(t) when |t – t*| < ε. Similarly,
we count the local minima at t*, if f(t*) f(t) when |t – t*| < ε.
RL for actor i over time window ws is therefore:
RLi = #local minimai
ws + #local maximai
ws
𝑅𝐿 =
1
𝑛
𝑅𝐿!
!
!!!
Figure 1: RL visualized through oscillations in BC over time [17]
This figure illustrates Rotating Leadership (RL) for two teams, one with high RL, and one with low RL.
Actors are placed along the Y-axis, while the X-axis encodes time, and the Z-axis the Betweenness
Centrality (BC) of actors for each hour, sorted, each hour, by the decreasing BC of actors. The back
plane, which rises and falls, represents the set of actors who rotate taking the lead in the team’s
communication.
The left picture illustrates an example of a team with high RL. This example was drawn from a 6-day long
graduate student seminar, and communications were measured using sociometric badges. This image
includes 15 actors, and has had BC oscillation computed every hour using a time window of 12 hours,
with a date range 6/13/2010 12:37 pm to 6/19/2010 23:37 pm.
The right picture illustrates an example of a team with low RL. This example was drawn from the
customers and employees of a large service provider serving one customer, and communications were
measured using the email archive of the service provider. This image includes 2857 actors, and has had
BC oscillation computed every day using a time window of 7 days, with a date range between 6/13/2012
to 12/30/2012. The high back represents the key account managers who are consistently taking the lead in
team communication.
(2) Rotating Contribution (RC)
Rotating Contribution (RC) measures the degree to which, over time, actors in a team vary in how much
they broadcast communications versus listen to communications. RC can be observed in a visualization of
a network when distinct nodes appear, over time, to vary widely in how many incoming versus outgoing
links they have. Intuitively, RC evaluates how much, across time and the team members, team members
switch off between broadcasting many communications and listening to may communications. Formally,
RC measures oscillations, over time, of the Contribution Index (CI) of actors in a team.
Contribution Index (CI) is a measure of how much an actor disseminates versus receives communications.
For a given node, it is equal to ratio of incoming to outgoing links incident upon that node. An actor that
only sends messages will have a CI of 1, an actor that sends and receives an identical number of messages
will have a CI of 0, and an actor that only receives messages will have a CI of -1 [18]. Formally, the CI of
an actor over a given time frame is:
CI =messages _sent messages _received
messages _sent +messages _received
In order to calculate RC, it is necessary to aggregate measures of CI, which occur at the scale of an
individual actor at an individual time step, to the scale of the whole network over the full time frame. In
order to do this in a fashion that indicates variation in CI we counted the number of local maxima and
minima in the vector of each actor’s CI over time, and then summed this number across the actors in a
team.
Formally, we count the number of local maximum points of function f(t)=c(t) within time interval [t1,t2].
There is a local maximum for time t at point t* if there exists some ε > 0 such that f(t*) f(t) when |t – t*|
< ε. Similarly, we count the local minima at t*, if f(t*) f(t) when |t – t*| < ε. RCi
ws for actor i and time
window ws is therefore
RCi
ws = #local minimai
ws + #local maximai
ws
RC= !
!
𝑅𝐶!
!
!!!
Figure 2: RC visualizing though CI oscillations over time [17]
This figure illustrates Rotating Contribution (RC) for two teams, one with high RC, and one with low RC.
Actors are placed along the Y-axis, while the X-axis encodes time, and the Z-axis the Contribution Index
(CI) of actors for each hour, sorted, each hour, by the decreasing CI of actors. The back plane, which
rises and falls, represents the set of actors who rotate taking the lead as most vocal contributors.
The left picture illustrates an example of a team with high RC; RC oscillates highly among time steps and
the actors of the team. This example was drawn from a 6-day long graduate student seminar, and
communications were measured using sociometric badges. This image includes 15 actors, and has had
BC oscillation computed every hour using a time window of 12 hours, with a date range 6/13/2010 12:37
pm to 6/19/2010 23:37 pm.
The right picture illustrates an example of a team with low RC; CI oscillates relatively little among time
steps and the actors of the team. This example was drawn from the customers and employees of a large
service provider serving one customer, and communications were measured using the email archive of
the service provider. This image includes 2857 actors, and has had CI oscillation computed every day
using a time window of 7 days, with a date range between 6/13/2012 to 12/30/2012. The high back
represents the key account managers who are consistently the most vocal by sending more emails than
they receive.
(3) Prompt Response Time (PRT)
Prompt Response Time (PRT) measures the degree to which, over time, actors are prompt at
communicating to those who have communicated to them. Intuitively, PRT corresponds with how fast,
across actors in a network, actors are at “getting back” to each other’s communications. In order to
capture this formally, PRT is defined in terms of the Communication Frame (CF), which groups
communication events between pairs of actors which may “get back” to each other, and Frame Nudges,
which measure the number of communication events in a CF, and Elapsed Time, which measures the time
duration of a CF.
A Communication Frame (CF) groups a set of time-adjacent communications between a pair of actors.
Suppose a pair of actors X and Y in a network, with a set of communication events, or directional, time-
stamped edges, between them. A single CF defines all communication events from X to Y, prior to and
including a communication event from Y to X. Intuitively, this is all the messages your colleague has sent
you since you last messaged her, prior to and including the first message you send back to your colleague.
In this framework, you, actor X, are the “source” actor in the CF, corresponding with the origin of the
first communication in the CF, and your colleague, actor Y, is the “target” actor in the CF, corresponding
with the origin of the last communication in the CF. The Elapsed Time (ET) for this CF is the difference
between the first communication in the frame and the last communication in the frame. The Frame
Nudges (FN) for this CF is the number of communication events in the CF, intuitively FN is the number
of “pings” X sends until Y responds.
To get the network-level measure of PRT from the edge-level measure of ET in CFs it is necessary to
aggregate measures. We accomplished this by using an intermediate actor-scale measure, where we
evaluated the “responsiveness” of actors through their Responsiveness in Communication Frames (RCF).
Intuitively, we measure how quickly actors got back to people who messaged them.
This can be accomplished either by measuring the ET or the FN of CFs. We define RCF via ET (RCF-
ET) for an actor as the mean ET for all CFs in which the node is the “target” node. We define RFC via
FN (RCF-FN) for an actor as the mean FN for all CFs in which the node is the “target” node. For actor i,
where 𝑓 is a given CF in the set of CF denoted as 𝑓F, and T
!!is the time elapsed for frame 𝑓, such
that 𝑓 Fi, where Fi represents the set of all frames that actor i is a member, RCF -ET is:
RCF-ET i =
!!
|!!|
!!!
!!
For actor i, where 𝑓 is a given CF in the set of CF denoted as 𝑓F, and FN!is the number of edges in
frame 𝑓, such that 𝑓 Fi, where Fi represents the set of all frames that actor i is a member, RFC-FN
is:
RCF-FN i =
!!!
|!!|
!!!
!!
We then aggregate this actor-scale measure to the networks-scale by averaging RCF for all actors in the
network. This procedure is the same for RCF-ET and RCF-FN. For a network, where 𝑛! denotes the
number of communications of each actor i, PRT is therefore:
PRT= !"#!!!!!
!
!!!
!!
!
!!!
!
Analysis and Results
We extracted signals of team creativity and productivity from electronic records of interpersonal
interactions, including e-mail, and face-to-face interaction measured via sociometric badges [28]. Some of
our samples have quite a small N (<10) because of the difficulty of obtaining the type of small group
communication data we are analyzing - small team communication networks which are associated with a
measure of creativity and/or performance. This is compensated for by the comparability of the 5 datasets
that allow for cross-comparative validation across a wide range of small group and larger organizational
settings. For each scenario, we measured the Rotating Leadership, Rotating Contribution, and Prompt
Response Time measured by Elapsed Time and Frame Nudges, for the teams recorded (Table 1). We
studied the following five scenarios, captured via the described datasets:
Dependent
variable
Interaction
type
#interaction
records
duration
Global Virtual
Course
creativity
e-mail
3782
3 months
Co-located Course
creativity
sociometric
badges
265,160
5 days
Eclipse developers
Creativity,
performance
mailing list
6405
6 months
Medical
researchers
creativity
e-mail
117,027
12 months
Service Provider
performance
e-mail
7,640,016
7 months
Table 1. Basic parameters of 5 datasets employed to verify “honest signals”
(1) (COINscourse2012 [19]) An e-mail archive of a multinational, distributed graduate student seminar.
Contains 161 actors and 3782 messages. 50 students were divided into 10 student teams at five in three
countries on two continents. These students worked together as distributed virtual teams over 14 weeks.
The dependent variable for creativity was taken as the mean of peer-ratings of students, and from an
instructor rating.
(2) (CGSseminar2010 [20]) A face-to-face interaction archive of a co-located course, gathered through
sociometric badges at a doctoral seminar. Contains 15 participants, who worked on different projects in
nine teams during one week. The dependent variable for creativity was measured through peer ratings
from participants.
(3) (Eclipse2005 [12]) - A mailing list archive of 26 working groups of Eclipse open source developers.
Contains 1371 actors and 6405 messages over a period of six months. The dependent variables for
performance and creativity were measured as the volume of bugs fixed (normalized by team size) and the
volume of new features (normalized by of the count of fixed bugs), respectively.
(4) (ChronicCareTeams2012 [21]) An e-mail archive with a core team of 30 clinicians and health
services researchers. Contains 22,523 different actors and 117,027 messages, working on 10 different
medical innovations over the period of one year. The dependent variable for creativity was measured
through ratings from a senior project management team.
(5) (ServiceProvider2012 [22]) An e-mail archive of staff members working in 14 different accounts at
a global service provider. Contains 85,680 actors and 7,640,016 different messages, with an account
manager coordinating activities per corporate customer. The dependent variable for customer satisfaction
was measured through a Net Promoter Score [23].
COINscourse2012
RL
RC
PRT - FN
PRT - ET
creativity
Pearson Correlation
.830**
.928**
.796**
-.610
Sig. (2-tailed)
.003
.000
.006
.061
N
10
10
10
10
CGSseminar2010
creativity
Pearson Correlation
.707*
.733*
.368
.275
Sig. (2-tailed)
.033
.025
.370
.509
N
9
9
8
8
quality
Pearson Correlation
.277
.261
.882**
.954**
Sig. (2-tailed)
.470
.498
.004
.000
N
9
9
8
8
Eclipse2005
bugs_fixed
Pearson Correlation
-.092
-.200
-.366
-.546**
Sig. (2-tailed)
.654
.328
.078
.006
N
26
26
24
24
performance
Pearson Correlation
-.754**
-.698**
-.266
-.161
Sig. (2-tailed)
.000
.000
.220
.462
N
25
25
23
23
creativity
Pearson Correlation
.216
.246
.554**
-.084
Sig. (2-tailed)
.289
.226
.005
.697
N
26
26
24
24
ChronicCareTeams2012
creative
performance
Pearson Correlation
.753*
.751*
-.117
.262
Sig. (2-tailed)
.012
.012
.749
.465
N
10
10
10
10
creativity
Pearson Correlation
.231
.287
.730*
.571
Sig. (2-tailed)
.520
.422
.017
.085
N
10
10
10
10
ServiceProvider2012
performance
Pearson Correlation
-.589*
-.618*
.429
.629*
Sig. (2-tailed)
.034
.019
.164
.029
N
13
14
12
12
Table 2. Correlations between 3 Social-Network based indicators and Creativity for the five test
datasets. **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05
level (2-tailed)
Discussion
Dynamic Social Network Analysis (SNA) [24] provided us a common framework across these scenarios,
allowing us to extract the same measures. SNA represents people as nodes and their connections as links
which together form a network. The properties of the resulting network and its entities can be studied to
glean insights about the human collection represented. SNA has been previously used to study creativity
[25][26]. While adding time at the actor level is not new [27], our work complements existing methods by
measuring interaction over time among teams of individuals who must necessarily communicate,
allowing us to measure edge-dependent features of the network as well.
A main limitation of our study is the small N (<10) of some of our samples. This is caused by the
substantial effort of obtaining the type of small group communication data we are analyzing - team
communication networks which are associated with measures for creativity and performance. This
limitation is compensated for by the comparability of the 5 datasets, allowing for cross-comparative
validation across a wide range of small group and larger organizational settings. We also hope that the
far-reaching insights into human creativity possible through the approach proposed in this paper will
motivate other researchers to conduct similar studies, thus increasing the availability of data to validate
and extend our approach.
Rotating leadership RL and Rotating Contribution RC are a consistent indicator of creativity, we also find
that for a non-creative activity such as large account management at the global service provider RL and
RC maintain predictive power, however the direction of the correlation changes: for creative tasks, more
is better, for non-creative tasks, less rotation in leadership and contribution is better. The number of
nudges PRT-FN is a predictor of high creativity, while counterintuitively taking more time (PRT-ET)
for a reply leads to more satisfied customers of the service provider. PRT-ET is positively correlated to
the speed of fixing software bugs, which makes intuitive sense: the faster developers answer, the faster
they will also be in fixing bugs.
While the results presented are preliminary, they nevertheless illustrate that “honest signals” of
communication among team members predict the creativity and performance of the team. They are
therefore a first step towards defining a new science of collaboration, that delivers a novel way to measure
and even optimize creativity and performance of teams by coming up with recommendations for
increased communication. While the definition of “creativity” remains elusive, we have introduced a set
of robust dependent metrics that have the power to predict if humans working together in a team might be
engaged in a creative task.
References
1. Wilson EO. (2012) The Social Conquest of Earth, Liveright Publishing Corp. New York. p. 17 and p 44.
2. Sawyer C. (2007) Group genius: the creative power of collaboration. Basic Books, New York.
3. Pentland A. (2008) Honest signals, How they shape our world. MIT Press, Cambridge
4. Gloor P, Oster D, Raz O, Pentland A, Schoder D. (2010) The virtual mirrorreflecting on your social and
psychological self to increase organizational creativity. Journal International Studies of Management and
Organization. 40(2) 74-94.
5. Gloor P, Fischbach K, Fuehres H, Lassenius C, Niinimäki T, Olguin Olguin D, Pentland A, Piri A, Putzke J.
(2010) Towards “Honest Signals” of Creativity Identifying Personality Characteristics Through Microscopic
Social Network Analysis. Procedia - Social and Behavioral Sciences, 26. Proceedings COINs 2010,
Collaborative Innovations Networks Conference, Savannah GA, Oct 7-9.
6. Csíkszentmihályi M. (1996) Creativity: flow and the psychology of discovery and invention. Harper Perennial,
New York.
7. Amabile TM. (1983) The social psychology of creativity. Springer-Verlag, New York.
8. Jung DI, Chow C, Wu A. (2003) The role of transformational leadership in enhancing organizational
innovation: Hypotheses and some preliminary findings, The Leadership Quarterly, 14(45), 525-544 August
October.
9. Woodman RW, Sawyer JE, Griffin RW. (1993) Toward a Theory of Organizational Creativity, The Academy of
Management Review, Vol. 18, No. 2, pp. 293-321.
10. Nemoto K, Gloor P. Laubacher R. (2011) Social Capital Increases Efficiency of Collaboration among
Wikipedia Editors, ACM Hypertext 2011: 22nd ACM Conference on Hypertext and Hypermedia, Eindhoven,
NL, June 6-9.
11. Freeman L. (1977) A set of measures of centrality based on betweenness. Sociometry 40 3541.
12. Kidane Y. Gloor P. (2007) Correlating temporal communication patterns of the Eclipse open source community
with performance and creativity, Computational & Mathematical Organization Theory. 13 (1). 17-27.
13. Fischbach K, Gloor P. Schoder D. (2009) Analysis of informal communication networks A case study.
Business & Information Systems Engineering 2, February.
14. Olguin Olguin D. Gloor P. Pentland A. (2009) Capturing Individual and Group Behavior with Wearable
Sensors. Proc AAAI 2009 Spring Symposium, Stanford, March 23-25.
15. McCrae RR, Costa PT, Martin TA. (2005) The NEO-PI-3: A more readable revised NEO personality inventory.
Journal of Personality Assessment. 84 (3): 261270.
16. Woolley AW, Chabris CF, Pentland A, Hashmi N, Malone TW. (2010) Evidence for a Collective Intelligence
Factor in the Performance of Human Groups. Science. 29 October. 330 (6004) 686-688.
17. Gloor P. (2005) Capturing Team Dynamics Through Temporal Social Surfaces , Proceedings of 9th
International Conference on Information Visualisation IV05, London, 6-8 July.
18. Gloor P. Laubacher R. Dynes S. Zhao Y. (2003) Visualization of Communication Patterns in Collaborative
Innovation Networks: Analysis of some W3C working groups. ACM CIKM International Conference on
Information and Knowledge Management, New Orleans, Nov 3-8.
19. Gloor P, Paasivaara M. (2013) COINs Change Leaders - Lessons Learned from a Distributed Course.
Proceedings 4rd Intl. Conf on Collaborative Innovation Networks COINs 2013, Aug. 11-13, Santiago de Chile.
20. Gloor P, Grippa F, Putzke J, Lassenius C, Fuehres H, Fischbach K, Schoder D. (2013) Measuring Social Capital
in Creative Teams Through Sociometric Sensors. Int. J. Organisational Design and Engineering, 2 4.
21. Zhang X. Gloor P. Grippa F. (2013) Measuring Creative Performance of Teams Through Dynamic Semantic
Social Network Analysis. Int. J. Organisational Design and Engineering, 3 2.
22. Brunnberg D, Gloor P, Giacomelli G. (2013) Predicting Customer Satisfaction through (E-Mail) Network
Analysis: The Communication Score Card. Proceedings 4rd Intl. Conf on Collaborative Innovation Networks
COINs 2013, Aug. 11-13, Santiago de Chile.
23. Reichheld F. (2013) One Number You Need to Grow, Harvard Business Review. December.
24. Carley KM. (2003) Dynamic Network Analysis, in Dynamic Social Network Modeling and Analysis:
Workshop Summary and Papers, Ronald Breiger, Kathleen Carley, and Philippa Pattison, (Eds.) Committee on
Human Factors, National Research Council, National Research Council. 133145, Washington, DC.
25. Cummings JN, Kiesler S. (2008) Who collaborates successfully? Prior experience reduces collaboration barriers
in distributed interdisciplinary research. Proceedings of the ACM conference on Computer-Supported
Cooperative Work, (November 10-12), San Diego, CA (2008).
26. Leenders RT, van Engelen JM, Kratzer J. (2003) Virtuality, communication, and new product team creativity: a
social network perspective, Journal of Engineering and Technology Management, Volume 20, Issues 12, June,
69-92.
27. Snijders TAB. (2005) Models for Longitudinal Network Data. Chapter 11 in P. Carrington, J. Scott, & S.
Wasserman (Eds.), Models and methods in social network analysis. New York: Cambridge University Press.
28. Olguín Olguín D. (2007) Sociometric badges: wearable technology for measuring human behavior. MAS
Thesis, Massachusetts Institute of Technology. Cambridge, MA, May.
... While early work for data-driven modeling on collaboration behavior patterns has mainly aimed to model lower-level behavioral dimensions, such as turntaking [31], recent efforts go beyond low-level signals to model high-level collaborative behavioral patterns. For example, postural markers have been used in human activity recognition to differentiate team member group functions [6], and proxemic features have been shown to be indicators of knowledge-sharing dynamics and affect group creativity [17,12]. ...
... This is particularly applicable to colocated collaboration settings because face-to-face teamwork remains the dominant mode for solving complex problems despite the increase of virtual teams. Furthermore, colocated collaboration provides unique benefits that are not easy to achieve in digitally mediated forms of teamwork [28], such as increasing creativity [12] and performance [25]. While preliminary work has demonstrated the feasibility and utility of leveraging multimodal signals to predict behavioral patterns during collaboration activities, more research is needed to understand which data sources work best to predict certain activities. ...
Chapter
Full-text available
Despite the importance of team communication for successful collaborative problem solving, automated solutions for teams are notably absent from the literature. One promising avenue of research has been the development and integration of speech-based technology for team meetings. However, these technologies often fall short of meeting the needs of the teams as they do not take meeting context into consideration. In this paper, we demonstrate the efficacy of context detection with data collected during real team meetings. By capturing and analyzing social signals of rotation in team dynamics, we can demonstrate that different stages of collaborative problem solving using the design thinking methodology differ in their dynamics. Using supervised machine learning, we successfully predict design thinking mode with an overall F1 score of 0.68 and a best-performing sub-class model of 0.94. We believe this to be an essential step towards improving speech-based technology that aims to assist teams during meetings. Making these automated systems context-aware will enable them to provide teams with relevant information, such as resources or guidance. KeywordsSocial signalsContext detectionPredictive modeling
... actor to respond to the email messages he/she receives, or to tweets which are directed at him (Ego ART). A second measure -Alter ART -measures the time it takes for others to respond to a user or to tweets where a user is mentioned (Gloor, Almozlino, Inbar, Lo, & Provost, 2014). ...
... Nudges. This variable counts, on average, the number of pings (nudges) required before a social actor answers to an email or to a tweet directed to him (Gloor et al., 2014). Nudges can be subsequent emails sent to an employee who has not yet responded, or new tweets which keep mentioning a Twitter user before receiving an answer. ...
Preprint
Full-text available
This paper investigates the research question if senders of large amounts of irrelevant or unsolicited information - commonly called "spammers" - distort the network structure of social networks. Two large social networks are analyzed, the first extracted from the Twitter discourse about a big telecommunication company, and the second obtained from three years of email communication of 200 managers working for a large multinational company. This work compares network robustness and the stability of centrality and interaction metrics, as well as the use of language, after removing spammers and the most and least connected nodes. The results show that spammers do not significantly alter the structure of the information-carrying network, for most of the social indicators. The authors additionally investigate the correlation between e-mail subject line and content by tracking language sentiment, emotionality, and complexity, addressing the cases where collecting email bodies is not permitted for privacy reasons. The findings extend the research about robustness and stability of social networks metrics, after the application of graph simplification strategies. The results have practical implication for network analysts and for those company managers who rely on network analytics (applied to company emails and social media data) to support their decision-making processes.
... employee to answer to forum posts directed to him/her (Ego ART); we also considered the symmetrical measure -the Alter ART -which represents the time taken by the other users to answer to his/her messages (Gloor, Almozlino, Inbar, Lo, & Provost, 2014). ...
... Nudges. This measure counts the average number of pings, i.e. nudges required before a social actor answers to a message (Gloor et al., 2014). Nudges can be subsequent messages sent to a user which has not yet answered. ...
Preprint
Full-text available
In this study, we tested the robustness of three communication networks extracted from the online forums included in the intranet platforms of three large companies. For each company we analyzed the communication among employees both in terms of network structure and content (language used). Over a period of eight months, we analyzed more than 52,000 messages posted by approximately 12,000 employees. Specifically, we tested the network robustness and the stability of a set of structural and semantic metrics, while applying several different node removal strategies. We removed the forum moderators, the spammers, the overly connected nodes and the nodes lying at the network periphery, also testing different combinations of these selections. Results indicate that removing spammers and very peripheral nodes can be a relatively low impact strategy in this context; accordingly, it could be used to clean the noise generated by these types of social actor and to reduce the computation complexity of the analysis. On the other hand, the removal of moderators seems to have a significant impact on the network connectivity and the shared content. The most affected variables are closeness centrality and contribution index. We also found that the removal of overly connected nodes can significantly change the network structure. Lastly, we compared the behavior of moderators with the other users, finding distinctive characteristics by which moderators can be identified when their list is unknown. Our findings can help online community managers to understand the role of moderators within intranet forums and can be useful for social network analysts who are interested in evaluating the effects of graph simplification techniques.
... Finding the optimal level of group-creativity may require plasticity-rigidity alterations of social groups. In agreement with this assumption, an intermediate amount of long-range connections (Guimera et al, 2005;Shore et al, 2015) resulting in the simultaneous presence of boundary spanning brokerage and trust-building closure (Tortoriello & Krackhardt, 2010;Uribe & Wang, 2014), as well as rotating leadership and contribution (Gloor et al, 2014) were shown to be key factors of teamsuccess in business, arts, sports and science. Changes of group-plasticity and rigidity dominance may be an important learning mechanism of social groups as detailed below. ...
Preprint
Network support is a key success factor for talented people. As an example, the Hungarian Talent Support Network involves close to 1500 Talent Points and more than 200,000 people. This network started the Hungarian Templeton Program identifying and helping 315 exceptional cognitive talents. This network is a part of the European Talent Support Network initiated by the European Council for High Ability involving more than 300 organizations in over 30 countries in Europe and extending in other continents. These networks are giving good examples that talented people often occupy a central, but highly dynamic position in social networks. The involvement of such 'creative nodes' in network-related decision making processes is vital, especially in novel environmental challenges. Such adaptive/learning responses characterize a large variety of complex systems from proteins, through brains to society. It is crucial for talent support programs to use these networking and learning processes to increase their efficiency further.
... Numerous studies in various domains have attempted to explain what makes groups work together effectively. Approaches in the area of human-computer interaction have focused on social signal processing to understand the impact of communication channels 2 and 3 on various team outcomes such as creativity [8] or performance [16]. However, previously published studies on the effect of social signals on team outcomes are not consistent, identifying various signals as potentially relevant. ...
Chapter
Full-text available
While meetings take up a significant part of the workday, participants often perceive them as poor and unproductive. With the surge in videoconferencing meetings for work due to the COVID-19 pandemic, many employees experienced that videoconferencing can even aggravate negative experiences in meetings. Past research has shown that the level of engagement during meetings is a crucial aspect of meeting a success. While there have been some attempts towards utilizing post-analysis feedback, there is little effort towards real-time support to improve engagement. This research explores the development of a visual support system for automated, real-time feedback on team communication behavior during online meetings. We present a novel, fully working visual support system that was evaluated with positive results. This study outlines the step-wise development of the method. We collected a range of qualitative feedback measures to understand better how users perceive the visual support system. First, we collected qualitative feedback from participants and eye-tracking data (n = 4) to evaluate four visualization approaches. The second step evaluated the best-performing visualization by a user study with participants (N = 72) working in groups of four on a collaborative problem-solving task. Users give the tool good scores on a seven-point Likert scale: perceived usefulness (4.8), ease of understanding (5.6), and perceived precision (5.1). Our results indicate that our novel system can enhance the quality of video conferencing through real-time visual support.
... In this study, we explore possible cues in the managers' communication behavior that indicate a change in the relationship "managers-organization" and possibly a fracture in the psychological contract. Following a method similar to the embeddedness approach to turnover (Mitchell, Holtom, Lee, Sablynski, & Erez, 2001) we used new social network metrics such as betweenness centrality oscillation, average response time, nudges and emotionality metrics (Allen, Gloor, Fronzetti Colladon, Woerner, & Raz, 2016;Gloor, Almozlino, Inbar, & Provost, 2014) to identify changes in the communication behaviors of managers who are close to quit their job. ...
Preprint
Full-text available
In this study we propose a method based on e-mail social network analysis to compare the communication behavior of managers who voluntarily quit their job and managers who decide to stay. Collecting 18 months of e-mail, we analyzed the communication behavior of 866 managers, out of which 111 left a large global service company. We compared differences in communication patterns by computing social network metrics, such as betweenness and closeness centrality, and content analysis indicators, such as emotionality and complexity of the language used. To study the emergence of managers' disengagement, we made a distinction based on the period of e-mail data examined. We observed communications during months 5 and 4 before managers left, and found significant variations in both their network structure and use of language. Results indicate that on average managers who quit had lower closeness centrality and less engaged conversations. In addition, managers who chose to quit tended to shift their communication behavior starting from 5 months before leaving, by increasing their degree and closeness centrality, the complexity of their language, as well as their oscillations in betweenness centrality and the number of "nudges" they need to send to peers before getting an answer.
Article
Last years one can observe a growing interest of researchers and practitioner in organizational social networks based on different forms of relations between employees. Such networks may refer to formal or informal relations between employees. The former of these relations may stem for example from the position of each person in an organization, and the latter may refer to relations between employees based for example on trust or confidence. Social network analysis tools provide several measures which allow to analyze different properties of such networks. The paper focuses on link prediction in organizational social network, where relations between employees are defined by intensity of digital communication (e-mail channel) between them. Different similarity measures have been considered by Authors. Basing on similarity scores calculated for all non-existing links in the network, new potential links with highest likelihood of their formation have been discovered. The results of the experiment (predicted links) have been compared with the results obtained by solving a community detection problem in the same network. They have been also compared with a real structure of the organization investigated in the experiment. The experiment confirmed that the results of solving a link prediction problem in the organizational network may support the process of detecting communities in the network, but it strongly depends on the similarity measure used.
Conference Paper
Full-text available
Collaborative creativity is an essential part of modern teamwork and is often supported by formal techniques, such as design thinking. Current support tools are often limited in scope as understanding the time-varying nature and structure of team communication is insufficient. We investigate how collaborative creative activities in new product development teams can be digitally supported while maintaining face-to-face communication. This work analyzes to what extent paralinguistic and proxemic features of team interaction relate to performance in new product development teams and if and how this relationship differs for different stages in the design process. This is investigated by applying multilevel modeling on data collected during a four-week new product development cycle. The cycle was completed by four teams, during which data were collected automatically using sociometric badges that capture social signals of team interactions. In addition, the data are combined with survey-based measurements on the team’s daily design process and periodic performance evaluations. The current paper provides evidence that social signals are related to team performance and that this relationship varies across the stages in the product design process. Certain social signals contribute positively in one stage but less in other stages, showing the importance of using multimodal signals when modeling high-level collaborative patterns. This research contributes to the literature by providing a better understanding of relevant factors when designing supporting tools or methods for collaborative creative problem solving.
Chapter
Nowadays a lot of diverse systems in many different fields can be described as complex network and they are the focus of interest in many disciplines such as politics, marketing, social systems. Using different network analysis tools may provide many interesting observations about the structure of the network, dynamics of the network over the time and the role of selected nodes in the network. The paper focuses on organizational social network based on email communication between employees within the organization. Such network has a form of a network including a set of vertices, referring to persons employed in this organization, and a set of edges, defining information flow between these persons using an email communication channel. The main contribution of the paper is to discover main properties of the email-based social network of public organization located in Poland and to identify key actors in it using social network analysis tools. An important part of the analysis is also a comparison of the obtained results with real structure of the organization. The experiment confirmed that analysis of email traffic within an organization may derive information that can be usable for organizational management purposes.
Article
Full-text available
This paper presents preliminary results on predicting individual creativity based on interpersonal interaction patterns. We combine insights from analyzing communication in an E-mail student network of a distributed course with measurements of interaction by sociometric badges for 23 programmers in Northern Europe. In the first case study we measure communication patterns of 23 software developers in a Nordic country through sociometric badges over a period of 4 weeks, associating it with creativity and productivity collected by a daily questionnaire. At the same time we collected individual trust through another questionnaire. We found that the more central people are in the network, the more trusting they are, and the less they oscillate between low and high states of energy, the more trusting they are. The second case study is based on a sample of 17 students from a German university participating in a multinational course. We show that e-mail behavior is associated with personality type as measured by the FFI personality test. We found that the larger degree and betweenness centrality of students in the e-mail course network is, the more agreeable and less neurotic they are. The faster students respond to e-mail, the more open and agreeable they are. The smaller the contribution index of students is, i.e. the less e-mails they send relative to other team members, the more neurotic they are.We speculate that there might be two different types of creativity, “lonely genius” - feeling most creative when on his/her own with lower trust in others, and “swarm creative” - most creative when in the midst of other people, and with higher trust.
Article
Full-text available
This study seeks to better understand the network characteristics of client support teams by analyzing the teams' e-mail communication networks and comparing it to client organization's satisfaction. In collaboration with a large service provider we studied the impact of network properties on the satisfaction of client organizations. In particular, we found that social network metrics correlate with client satisfaction as measured by Net Promoter Score (NPS). A Communication Score Card is suggested as a dashboard to continuously measure client satisfaction, illustrating that data-driven analysis might help improving service providers' service quality management.
Article
Full-text available
In this paper we analyze the communication network of 50 students from five universities in three countries participating in a joint course on Collaborative Innovation Networks (COINs). Students formed ten teams. Interaction variables calculated from the e-mail archive of individual team members predict the level of creativity of the team. Oscillating leadership, where members switch between central and peripheral roles is the best predictor of creativity, it is complemented by the variance in the amount of sending or receiving information, and by answering quickly, and positive language. We verify our automatically generated creativity metrics with interviews.
Article
Full-text available
A theoretical framework for understanding creativity in a complex social setting, such as an organization, is developed. Organizational creativity is defined as the creation of a valuable, useful new product, service, idea, procedure, or process by individuals working together in a complex social system. The starting point for the theoretical development is provided by the interactionist model of creative behavior developed by Woodman and Schoenfeldt (1989). This model and supporting literature on creative behavior and organizational innovation are used to develop an interactional framework for organizational creativity. The theoretical framework is summarized by 3 propositions that can effectively guide the development of testable hypotheses.
Article
We describe the results of an experiment capturing the face-to-face 'honest signals' of knowledge workers through sociometric badges. We find that collective creativity of teams is a function of the aggregated social capital of members. The higher it is, the higher their creative output. We collected communication data of 14 graduate students and their instructor during a one-week seminar, comparing it against the creative output of their teamwork. As a second component of social capital we also measured the level of trust team members show to each other through surveys. We find that the more team members directly interact with each other face-to-face, and the more they trust other team members, the more creative and of higher quality the result of their teamwork is.
Article
In this project we compare communication structure and content exchanged by members of creative, interdisciplinary teams of medical researchers, physicians, patients and caretakers with their creative output. We find that longitudinal social networking patterns and word usage predict creative performance. We collected the e-mail archives of 60 members of a community of researchers working on 12 projects improving various aspects of the daily lives of patients with Crohn's disease. Our results indicate that more creative projects show a decrease in group density, while more actors are involved, and more e-mails are exchanged, suggesting that a more successful project attracts more attention from many different people. We also found that members of more creative projects use more outspoken language, which gets more focused over time.
Article
This article offers a novel approach for managers to increase creativity in their teams by measuring individual personality characteristics, crucial for creative people. We assess inter-personal interaction with sensor-equipped badges worn on the body. In a research project with 22 study subjects, who wore the badges during work for one month, we were able to predict extroversion, neuroticism, openness, and agreeability based on microscopic social network analysis. We obtained control measures of these values with a standard psychological test (NEO-FFI). As opposed to conventional personality tests, where people have to fill out lengthy questionnaires and surveys, our method offers an automated, and potentially more reliable way to assess these personality characteristics. Once these characteristics are considered, teams can be reshuffled and team membership changed for higher creativity.