Content uploaded by Poornima Kundapur
Author content
All content in this area was uploaded by Poornima Kundapur on Jul 31, 2017
Content may be subject to copyright.
Abstract—The core research issue presented in this article is
to study the factors that influence the knowledge worker’s
adoption of Knowledge Management System (KMS). The five-
step System Dynamics research methodology was followed to
design and develop the Knowledge worker adoption model.
This model studies the transition of a knowledge worker from
a non-user to an experienced user over a fixed time frame
identified. Simulation experiments were conducted show that
knowledge workers respond differently as they integrate the
new technology into their work pattern. The results were
analyzed for the different employee categories based on their
level of expertise of usageof KMS in the organization and
factors such as supervision, ease of use of technology were
found to affect the knowledge worker’s concerns for adopting
KMS in their work. The current simulation model was built
using the softwareVensim® and the outcome of the study
should aid administrators and policy makers to evaluate the
impact ofthe identified factors on adoption of KMS.
Index Terms—Knowledge worker, knowledge management
system, system dynamics, KWAM.
I. INTRODUCTION
The impact of globalization and technological advances
continue to change the competitive business environments,
making knowledge and expertise primary sources to
leverage the competitive advantage, at least in knowledge-
intensive industries.
For most firms Knowledge Management (KM) is
achieved through a series of initiatives that seek to build a
culture and infrastructure that connects people and
processes [1]. However, in the present competitive business
settings, the manner in which organizations learn from past
performances and manage knowledge through their most
important tangible asset, the knowledge worker force, has a
huge impact on future decisions.
The role of technology or that of knowledge strategy in
an organization depends not only on the knowledge
infrastructure of a company, but also on the attitude of
knowledge workers towards knowledge sharing, creation
and use of technology and towards the technology itself.
Once implemented in an organization, the success of KMS
implementation is determined based on the use and
acceptance of the system by knowledge workers.
One way to better understand the factors underlying the
acceptance and behavioural patterns of knowledge workers
in an organization may be by applying the simulation
approach. This paper attempts to understand this aspect of
an organizational knowledge worker.
Manuscript received April 20, 2012; revised May 15, 2012.
The authors are with the Manipal Institute of, Manipal, Karnataka
576104 INDIA (e-mail: poornim.girish@ manipal.edu;
rodrigusr@rediffmail.com).
A knowledge worker’s concern of adopting a new
technologylike KMSwhen disseminated into any intellectual
marketplace such as the software sector or an IT company
or any other knowledge based organization, can display a
wide variety of behaviour. A number of dynamic variables
play an important part in the successful adoption of KMS by
the knowledge worker. It is seen that knowledge workers
often may have quite a few concerns as they adopt the new
technology, including factors like their individual training in
technical skill sets, academic background. The more
concerns they have, the more likely they are to be resistant
in adoptingthe system. Thus, it becomes imperative to
identify the factors that can affect the knowledge worker’s
adoption behaviour.
There is an inherent difficulty of testing such variables in
real time scenarios mainly due to the cost of conducting
such experiments. There is also the issue of knowledge
workers unwilling to share these issues. In such cases,
simulating the adoption process is a viable option that will
provide trainers and decision makers with methods to assess
the factors that support new technology use in any
organization. Hence, the research methodology applied in
this paper follows the principles of System Dynamics
method (SD), first introduced by J. W. Forrester in the
1960s at the Massachusetts Institute of Technology (MIT),
Boston[2]. The SD approach includes Problem
identification, System conceptualization, Model formulation,
Simulationand validation and Policy analysis and
improvement.The stock and flow modelling and simulation
are performed usingVENSIM PLE® software.Simulation
models generate behaviour throughsimulation.The SD
process is iterative and flexible [3].
II. OBJECTIVES
The main purpose of this research article is to propose a
simulation model that tests the impact of factors that affect
knowledge workers’ adoption of KMS. To achieve this
purpose, the following objectives have been formulated:
III. LITERATURE REVIEW
A. KM, KMS and Knowledge Worker
Knowledge Management is about creating, storing,
accessing and reusing knowledge to accomplish
organizational goals. In other words, Knowledge
management (KM) is the process of identifying and
Simulating Knowledge Worker Adoption Rate of KMS:
An Organizational Perspective
Poornima P. Kundapur and Lewlyn L. R. Rodrigues
International
Journal of Innovation, Management and Technology, Vol. 3, No. 4, August 2012
459
zIdentifying and relating variables within the system
zConstructing the Stock and flow diagrams
zFormulating the governing equations
zModeling and simulation of the Knowledge
Worker Adoption Model (KWAM)
leveraging the collective knowledge in an organization to
help the organization compete [4].
Alavi and Leidner [4] referred to KMS as an emerging
line of systems targeting professional and managerial
activities by focusing on creating, gathering, organizing and
disseminating an organization‘s ‘knowledge’ as opposed to
‘information’ or data.’
Any organization that has a KMS in place must require
individuals to develop, use and apply the organizational
knowledge. These individuals are termed “Knowledge
Workers” in today’s knowledge economy [5] [6].
B. The System Dynamics Approach
A KMS implementation would be based on a framework
that identifies with the working objectives of that
organization. In trying to understand how KM initiatives
work towards achieving organizational goals there is a need
to identify factors that influence knowledge workers’
acceptance of knowledge available in KMS and how these
factors in turn relate to the organizational environment.
Many KM initiatives and the KM literature have lacked a
theoretical foundation that can inform the process of KM
system development and in particular the process of KM
information systems development [1]. This aspect is where
we feel the system dynamics approach may help to facilitate
understanding and can enhance organizational KM practice.
Figure 2 illustrates the model that will be referenced in this
paper.
IV. METHODOLOGY
Fig. 1. Steps of system dynamics [8].
The research methodology adopted is in accordance to the
modeling process methodology as proposed by
Sterman3][8]. The steps as illustrated in the Fig 1 include:
C. Problem Articulation
The problem articulationis the initial and most important
step of the system dynamic approach that involves defining
the problem. In this paper, the problem identified is “To
identify the factorsthat affect a knowledge worker’s
adoption rate of a KMS and to establish a relationship
between the factors and the behavioural pattern”. According
to Sterman [3] a problem should never model a system,
because the problem determines which factorsare important
to include and which to exclude and therefore be used to
find the relevant system boundaries of the problem
A reference mode (the hypothesized behaviour of the
problem) and the time horizon of interest must be identified
in this stage[3][7].
D. Dynamic Hypothesis
This is the second step in the system dynamics modeling
process. Once the problem has been articulated and the
initial characterization is done, it is necessary to develop a
theory about the problem. This theory or hypothesis is
called “dynamic hypothesis”. The hypothesis is said to be
“dynamic” because it characterizes the dynamics involved
in the system to be modeled over the given time horizon. At
this point, feedback mechanisms and the delays involved in
the system are taken into account.
E. Formulation of a Simulation Model
The next step in the System dynamics methodology for
modelling is to move from the conceptual realm of diagrams
to a fully specified formal model, complete with equations,
parameters and initial conditions that can be simulated via
computerised software [3].
F. Testing of the Model
Sterman [3] explains that testing begins as soon as the
first equation is formulated. Testing partly involves
comparing the simulated behaviour of the system under
study with the actual behaviour. It also involves something
more where each equation is checked. Whether each
variable under consideration has a meaningful concept in
real time is also verified. Parametersensitivity checks are
useful to decide how much effort should be dedicated to
increasingthe precision of the parameters.
G. Policy Design and Evaluation
Once the structure and behaviour of the model have been
finalised and there is a certain amount of confidence in the
simulated model, the modeller can now move on to
designing and evaluating policies for improvement [7].
V. THE MODEL CONSTRUCTION
The model has been designed and developed based on the
generic basic diffusion model incorporated into the KMS
scenario of an organization [3][9][10].A knowledge
worker’s cycle of growth startsfrom a being a new
employee with no experience in using a KMS to a trainee
employee (undergoing training to use a KMS) to a new
knowledge worker (trained and ready to apply his skills in
using a knowledge repository like a KMS) to an
experienced knowledge worker with years of experience in
handling and applying KMS knowledge all treated as stock
variables in the model.
Causal loop diagrams are powerful tools to map feedback
structure of complex systems but they are limited by their
inability to show stocks and flows. Hence we have used
stock and flow diagrams to simulate the knowledge worker
behavioral pattern.
A. Stock and Flow Diagram
A knowledge worker in any organization utilizes the
KMS to capture contextual knowledge applicable to his area
of work. A knowledge worker‘s competence depends on his
understanding of his work profile together with the
information or knowledge seeking attitude he possesses [9].
However this is made possible if the KMS existing in the
International
Journal of Innovation, Management and Technology, Vol. 3, No. 4, August 2012
460
organization provides access to all the available knowledge
to the knowledge worker.
B. Stocks and Flows
In the system dynamics approach, causal loop diagrams
are well suited to represent interdependencies and feedback
processes however the drawback of using this tool is that it
has nothing to offer in terms of capturing stock and flow
structure of systems under study. Stocks and flows are an
essential concept in system dynamics theory [3] and their
importance lies in the
C. Identifying Stocks
The stock and flow diagram of the proposed KWAM is
presented in Fig. 2. The four stocks identified are:
New Employees: Indicating pool of employees who have
just joined the organization
Trainee employees: Indicating pool of employees who are
undergoing training
New Knowledge workers: Indicating pool of employees
who have completed training
Experienced Knowledge workers: Indicating stock of
employees who have been using the technology and
experienced users
D. The Adoption Process
All categories of knowledge workers whether they are
new employees, trainees or experiencedwork in the same
organization. The adoption process is about experienced
workers creating awareness among non-users about the use
of technology in their work. The model therefore focuses on
this rate of this interaction.The variable “knowledge worker
with non-user contacts” represents that pool of knowledge
workers who have adopted the systemcoming into contact
with employees who are non-users. Going ahead, there is a
reasonable chance that this contact may result in the non-
user adopting the system in future. “Adoption fraction”
represents this probability of conversion.
“Application fraction” is the model variable that refers to
the time fractionexperienced knowledge workers may
devote to skill application development including the time
they spend on doing research, publishing white papers or
working on resolving their project problems.The variable
“trainee conversions” is affected by any addition of trainee
employees who complete their training period.
Apart from the variables used, this model also includes
six constant values that determine the speed of transition of
knowledge workersr from the training phase to gaining
experience. The constant “Self-training time” represents the
time required for an employee with no formal training
tobecome sufficiently proficient to be a Knowledge worker
and “Minimum training time”indicates the time required for
a trained employee to become proficient in the technology
used. It is also observed that as the experienced knowledge
workerdevotesmore time to training, there is a change in the
average training time which moves from self- training time,
to minimum training time according to training productivity
change. Fig.2. depicts the Knowledge worker adoption
dynamics model.
Fig. 2. Stock and flow of knowledge worker adoption dynamics.
E. The Governing Equations
There are causal relations between the variables of the
model and these are linked in the form of equations for
quantifying the simulation results. The units of variables
are indicated in parentheses.
Knowledge workers =Experienced knowledge workers +
New knowledge workers (Units: users)
Experienced knowledge workers= INTEG (conversion
rate, initial workers) (Units: users)
Trainee employees= INTEG (adoption rate-trainee
conversion, 0) (Units: users)
New knowledge workers= INTEG (trainee conversion-
conversion rate, 0) (Units: users)
VI. ANALYSIS
A. Model Scope
The variables defined in the model are endogenous to the
model and serve the purpose specified as per the
requirements of SD model boundary identification.
B. Time Horizon
System dynamics may be used as a prediction tool, and
helps understand the problem being studied as well as the
potential decisions that may be considered. Hence the
modellr must be able to design for a particular purpose
outside a narrow time zone. In this case the trend of
technology life span averages around 10 to 15 years.
However the study maintains a 10 year time horizon at
TIME STEP=0.125[10].
C. Modeling Conditions and Results
This model is simulated at three extremes
New Employees Trainee employees New knowledge
workers Experience d
knowledge workers
adoption rate trainee conver sion conversion rate
contact rate
Non user contacts Knowledge workers
with non us er co ntact s
Total employees
Knowledge
worker ratio
Knowledge
workers init ial wo rk
e
adoption fraction
effect of quality on
adoption
effect of quality on
adoption lookup
norm adoption
fraction
Average Quality
New quality
initiat ives existing quality
self tr aining time
Minimum t rainin g
time training
productivity training fraction Ap plic atio n
frac tion supervision
frac tio n
self experience
time
Minimum
experience time
supervision
productivity
<New knowledge
workers>
<Experienced
knowledge workers>
International
Journal of Innovation, Management and Technology, Vol. 3, No. 4, August 2012
461
1) application fraction = 1(all effort is devoted to work on
the job and new employees must train themselves)
2) training fraction = 1(all effort is devoted to training
new employees)
3) supervision fraction=1 (all effort is devoted to
generation experienced knowledge workers)
The following section explains the various simulation
experiments conducted and the results established thereafter.
Simulation 1:Effect of change in knowledge worker
adoption behaviour varying 3 factors:
Application fraction =1, Training fraction=1 and
Supervision fraction=1
Fig. 3. Effect on knowledge workers.
In this simulation, the behaviour of the system is
experimented by adjusting the value of three factors. When
application fraction =1, the knowledge worker’s effort is
entirely devoted to work on the job and new knowledge
workers must train themselves. Setting Training fraction to
1 means all the effort is devoted training new employees
while setting Supervision fraction =1 implies all effort is
devoted to generating experienced knowledge workers.The
governing equations are:
Application fraction= INITIAL(1-supervision fraction-
training fraction) (Units: Dmnl)
Supervision fraction=0 (Units: Dmnl)
Training fraction=0 (Units: Dmnl)
In case of new knowledge workers (refer Fig. 4), there is
a steady rise when the training and application fraction are
changed. The inference could be that the new knowledge
worker has undergone sufficient training to utilise the KMS
and apply the same in independently resolving issues or
even researching on other areas of concern. The knowledge
worker pool however shows a steady rise with the right
amount of training leading to self-sufficiency in knowledge
to enable applying this knowledge in work.
The experienced knowledge worker on the other hand,
displays a gradual increase in the knowledge application
curve over the training curve implying that an experienced
knowledge worker makes adequate use of knowledge
acquired.
Fig. 4. Effect of simulation 1 on experienced knowledge workers.
Fig.4. shows this trend of experienced workers (in
numbers) when changes to parameters are applied (users in
the graph is the dimension for the stock experienced
knowledge workers)
With new knowledge workers, there is an interesting
trend as observed in Fig. 5. There will be a steep rise in the
number of newly trained knowledge workers adopting the
KMS technology with adequate training together with the
application of the newly knowledge acquired in solving
their work related issues. However, simulation shows the
trend tends to dip over the next 10 years and stabilises
towards the end of the time line indicating technology
obsoleteness attributes to adoption patterns. Interestingly the
supervision factor does not have any major impact on the
increase in the adoption rate of new knowledge workers.
Simulation 2:To study the effect on knowledge workers
when: New quality = 0.6 (quality initiatives to 6%)
Self-training and Minimum training time are set at
half their values.
Supervision fraction and training fraction set at 0.1
(At 10% of experienced worker’s time)
Fig. 5. Effect of simulation 1on knowledge workers.
Fig. 5 indicates that both ease of use and quality can have
a significant impact on the speed of diffusion of a
technology such as KMS in any organization.Additionally,
when self-training time and Minimum training time was
halved, the trend remains the same but the curvedoes show a
significant increase as is the case when the supervision
fraction and training fraction are balanced at 10% revealing
a steady upward increase in KMS usage behaviour.
Fig. 6. Effect of change in quality and self-training time on knowledge
workers.
VII. TESTING AND VALIDATION OF THE MODEL
Testing of a system dynamics model is carried out so as
to uncover errors and find out the model’s limitations. This
helps build confidence in the model for a modeler.
Sterman[3] states verification means checking the truth or
reality of the model whereas validation means to ensure that
the model supports the objective truth. He further adds that
no model can ever be verified and validated. The reason is
because no model is ever an exact representation of reality
or truth since it is based on many limiting assumptions.
Thus the model can be verified and validated only based on
a set of limiting assumptions. Also, though the model’s
Knowledge workers
400
300
200
100
0
33333333333333
222222222222222
111111111111111
2000 2004 2008 2012 2016 2020 2024 2028
Time (Year)
users
Kn owled ge w ork ers : All Sup erv ision
111111111
Kno wled ge wor ker s : All Tra ining
2222222222
Kn owled ge w ork ers : All App lica tion
33333333
Experienced kno wledge workers
400
300
200
100
0
33333333333333
222222222222222
111111111111111
2000 2004 2008 2012 2016 2020 2024 2028
Time (Year)
users
Experienced knowledge workers : All Supervision
11111111
Experienced knowledge workers : All Training
222222222
Experienced knowledge workers : All Application
3333333
New knowledge workers
60
45
30
15
0
3
3
3
3
3
33
3
3
3
3
333
2
2
2
2
2
222
2
2
2
2222
111111111111111
2000 2004 2008 2012 2016 2020 2024 2028
Time (Year)
users
New knowle dge w ork ers : All Sup ervisio n
11111111
New knowle dge w ork ers : All Tra ining
222222222
New knowle dge w ork ers : All App licatio n
3333333
Knowledge Workers Ease of Use
400
300
200
100
0
33333333333333
222222222222222
111
1
111111111 11
2000 2004 2008 2012 2016 2020 2024 2028
Time (Year)
users
Knowledge workers : At 10%
1111111111
Knowledge workers : Min Time and Training
22222222
Knowled
g
e workers : Maintain
Q
ualit
y
3
3
3
3
3
3
3
3
International
Journal of Innovation, Management and Technology, Vol. 3, No. 4, August 2012
462
validity cannot be proved, its falsity can surely be proved.
Forrester [2] substitutes the term “validity” by “significance”
and states that the validity should be judged by the model’s
suitability for a particular purpose.
Sterman and Forrester [3],[2] state various fundamental
tests that can be carried out. Testing enables discovering of
errors and limitations of the model and further sets the basis
for modifying the model accordingly. However it is
imperative to note that testing is not a process that is done at
the end only after the model developed. It is a continuous
and iterative process that starts right from the initial stages
of model building. During every stage of model
development, testing is either explicitly or implicitly carried
out. Subsequently the model is continuously improvised and
corrected based on the results and feedback from the tests
the model.
Listed below are the tests that were carried out on the
model as a part of the verification and validation process:
A. Face Validity
This is a test of consistency that answers the question
“Does the model structure look like the real system? The
model developed in this paper closely resembles the real
time scenario of a knowledge worker’s progression from a
novice employee to an experienced knowledge worker and
the corresponding level of expertise of usage of KMS.
B. Dimensional Consistency Test
This test is a Test of Suitability and deals with the
dimensional units of stocks, flows and variables in the
model. The simulation software performs the dimensional
consistency check. Wherever, lack of units for variables or
dimensional errors were found, they were suitably corrected
to ensure that the dimensional inconsistencies were removed.
C. Structure Verification Test
This is another test of suitability that ensures whether the
structure of the model is consistent with the relevant
knowledge about the knowledge worker adoption process.
The model development was based on the inputs gathered
after an in-depth us literature review and the model is based
on the generic growth process model. Every effort was
made to keep the model structure consistent with the
information collected.
D. Parameter Verification Test
The question “Do the numerical values of parameters
have real system equivalents?” needs to be justified to
successfully claim that the model clears this test of
consistency. In this model, the parameters correspond
conceptually and numerically to real life. All categories of
Knowledge workers have their numerical value measured as
number of users.
VIII. CONCLUSIONS AND IMPLICATIONS
The KWAM developed and simulated in this research
paper provides insight into understanding the behavioural
factors that affect the knowledge worker’s rate of adoption
of based on training, ease of use and quality parameters. The
work started with a set of written hypotheses and worked on
building the KWAM.The simulation results indicate
factors like supervision, training and user-friendly
technology favour adoption among novice workers. Asteady
rise in the number of new knowledge workers using the
organizational KMS when the training and application
fraction are set to a high of 1 was observed. There was a rise
in adoption of KMS even with increased quality initiatives
and ease of use. System dynamics provides methods for
validation of the model. The model is validated using Face
validity test, Dimensional consistency test and Parameter
Sensitivity test. The model gives a basis understanding
reality and action to work on this understanding, however to
establish more confidence, data and reality checks need to
be implemented which will be worked upon as the next
phase of research. Further, KM researches may also refer to
this model and explore dynamic structures not identified
based on specific situation mapping.
REFERENCES
[1] T. H. Davenport and L. Prusak, Working Knowledge, Harvard
Business School Press, 1998
[2] J. W. Forrestor, “System dynamics, systems thinking and soft OR,”
System Dynamics Review, vol. 10, no. 2-3, pp. 245-256, 1994.
[3] S. John. 2000. Business Dynamics: Systems Thinking and Modeling
for a Complex World. Irwin McGraw-Hill.
[4] M. Alavi and D.E. Leidner, “Review: Knowledge management and
knowledge management systems: Conceptual foundations and
research issues,” MIS Quarterly, vol. 25, no. 1, pp. 107-136, 2001.
[5] N. Ikujiro and H. Takeuchi, “The knowledge creating company,”
Oxford, University Press, 1995.
[6] R. L. Chase, “Knowledge management benchmarks,” the Journal of
Knowledge Management, vol. 1, no. 1, September 1997.
[7] K-O. Vogstad, “A system dynamics analysis of the Nordic electricity
market: The transition from fossil fuelled toward a renewable supply
within a liberalised electricity market,” PhD thesis Norwegian
University of Science and Technology Faculty of Information science,
Technology, Mathematics and Electrical Engineering Department of
Electrical Power Engineering, December 2004 NTNU, ISBN 82-471
6890-1 (elektronisk) ISBN 82-471 6892-8 (trykt)
[8] P. P. Kundapur and L. L. R. Rodrigues, “System dynamics approach
to simulate KMS success model”, Global Journal of Management
and Business Research, vol. 10, issue 2 (Ver 1.0), April 2010
[9] B. J. Ranganath and L. L. R. Rodrigues, ‘System dynamics: Theory
and case studies,” I. K. International. 2008.
[10] W-T. Wang and C-Yeuh, “The application of the technology
acceptance model: A new way to evaluate information system
success,” System Dynamics Conference Proceedings, 2005.
Poornima P. Kundapur is an Assistant Professor,
Department of MCA, Manipal Institute of Technology,
Manipal University, Manipal -576104.
Her research interests include Knowledge
Management and System Dynamics. Her present
research is a trans-disciplinary work on Knowledge
Managements Systems and System Dynamics,
focusing on the factors that dynamically influence
usage of KMS in any organization.
Humanities and Social Sciences,Professor, Dept of
Mechanical and Manufacturing Engineering., at
Manipal Institute of Technology, Manipal. He is a
Ph.D. in System Dynamics.
His other areas of research include Total Quality
Management, Knowledge/ Technology Management,
Manufacturing and Human Resource Management.
International
Journal of Innovation, Management and Technology, Vol. 3, No. 4, August 2012
463
Dr. Lewlyn L. R. Rodrigues is the head of Dept. of