Individual Learning Style and Systems Tool Preferences
Dr. Sheldon Friedman
St. Joseph College
1678 Asylum Avenue
West Hartford, CT. 06117
(860) 231-5289
Email: Sfriedman@sjc.edu
Steven A. Cavaleri Ph.D.
Professor
Department of Management & Organization
Central Connecticut State University
1615 Stanley Street
New Britain, CT 06050
Tel (860) 832-3283
E-mail: 'cavaleri@ccsu.edu!
Mike Raphael Ph.D.
Department of Management & Organization
Central Connecticut State University
1615 Stanley Street
New Britain, CT 06050
Tel (860) 832-3283
Learning Styles and Modeling
Abstract:
Research was caried out in order to determine if a relationship existed between the use
of specific system thinking tools and Kolb leaming styles. Subjects were university
students who played the B & B Enterprise game, took the Kolb Leaming Styles Inventory
and responded to a questionnaire. These questions related to both the frequency and
preferences of use of specified system thinking tools and leaming styles. The results
revealed that for some of the leaming style orientations a significant relationship with use
and preferences of system thinking tools existed. These results indicate that further study
is needed to clarify any differences in relationships between groups and between
frequency and preferences. It is thought that if a set of relationships can be identified,
such knowledge could facilitate the introduction of systems thinking and system
dynamics to selected groups within an organization.
Keywords [Systems thinking, system dynamics, Kolb Leaming Styles]
Introduction
Typically, there is not a clearly known path to improving people's systems thinking
abilities or their competency in using systems thinking tools to perform analyses of
dynamic problems. Conventional training and educational approaches ignore individual
factors that may influence the rate at which someone may master systems thinking, such
as action leaming style and preference for using certain types of systems thinking tools.
This study examined the extent to which frequency of use of certain systems thinking
tools, and individual leaming styles were correlated and the extent to which they
influence performance on a management flight simulator.
The conventional method of helping people to become better at systems thinking is to
teach them to use various systems thinking tools and expose them to simulated
environments, such as those found in Microworlds (Kim, 1994). While there is some
evidence that participation in such training can be useful, this approach often fails to
consider the leaming styles or cognitive limitations of training participants. Sterman
(1994) has argued that effective leaming depends on making correct interpretations of
dynamic feedback and argues that such misinterpretations may be the causes of barriers
to leaming It has been suggested that one way to improve leaming is the use of computer
simulations, such as Management Flight Simulators, and designing formal system
models. Such formal models have advantages over either informal or mental models,
especially when it comes to leaming. These advantages are- formal models are explicit
and communicable. System dynamics models are capable of transparently exposing
assumptions being held by modelers about the causal reasons for underperformance.
They can also be subject to criticism, experimentation, and reformulation. Formal models
have the added benefit of being able to handle complexity fairly easily (Pugh and
Richardson, 1981). Simulations have the advantage of being able to present numerous
variables for evaluation-something that the human mind cannot easily do. The use of
simulations is based on the expectation that, during the use of a flight simulator, users
will have an integrative experience of how the system operates to cause its behavior.
Additionally, there may be greater awareness of the interdependencies among parts of an.
organization that would not otherwise be possible with traditional teaching methods
(Graham, 1990). The use of Microworlds as a leaming tool became popularized after the
publication of Senge’ s (1990), acclaimed book, The Fifth Discipline. The effect of
Microworlds on the ability of people to leam systems thinking has previously been
explored by researchers, such as (Morecroft and Sterman, 1994). Further investigations
aimed at assessing the impact of Microwords on systems thinking competencies have
been reported by Huz, Anderson, Richardson and Boothroyd (1997), Doyle (1997),
Durham (2002, 2003), Cavaleri and Sterman, (1997), Cavaleri (2002) and Friedman and
Cavaleri (2003). Such studies suggest that there are still, as yet unexplained,
complexities found in those cognitive processes inherent in leaming systems thinking and
in fommulating policy in complex dynamical social systems.
Richmond (1993) defines systems thinking in terms of competencies, “The systems
thinkers’ forte is interdependence. Their specialty is understanding the dynamics
generated by systems composed of closed-loop relations. Systems thinkers use
diagramming languages to visually depict the feedback structures of these systems. They
then use simulation to play out the associated dynamics.” (p.113) He also identifies seven
core systems thinking skills including: 1. dynamic thinking, 2. closed-loop thinking, 3.
dynamic thinking, 4. structural thinking, 5. operational thinking, 6. continuum thinking,
and 7. scientific thinking.
This exploratory study considers how such cognitive and leaming processes influences
frequency of use and preference for using certain systems thinking tools. According to
Kim (1994) “There are at least ten distinct types of systems thinking tools... They fall
under four broad categories: brainstorming tools, dynamic thinking tools, structural
thinking tools, and computer-based tools.” (p. 10) While there are other systems thinking
tools besides those identified by Kim, this research is delimited by the fact that the ten
types of tools identified by Kim were the focus of the systems thinking training explored
inthis research. Students were asked to rate the frequency of use of tools and their
perceived value toward achieving sustained profitability in a management flight simulator
experience.
Limits to Growth in Learning
Conventionally, the term Limits to Growth is used to describe a system archetype in
which constraints on capacity restrict the potential of systems to grow at a desired rate.
However, this term can also be used to describe the effect of constrained leaming
capacity on the growth of one’s systems thinking capabilities. The notion of cognitive
limits are quite popular in system dynamics literature, as expressed by concepts such as
Simon's (1957) bounded rationality and Senge (1990), who used the term learning
disability to describe the difficulties experienced by individuals and organizations, in
making sense of feedback gained from experience in ways that yield valid conclusions
about how causal pattems are likely to work in practice in the fubure. While we concur
that the aforementioned concepts are important contributions to the field of system
dynamics, we propose there is also another class of limit to leaming that stems from the
inherent leaming styles of individuals. There has been much written on the effects of
action leaming styles ranging from David Kolb’s seminal work to Reg Revans writing in
the 1980s, to Cavaleri and Seivert’s (2005) concept of knowledge bias. The concept of
knowledge bias describes tendencies or pattems of individuals to leam from experience
to ways that favor certain parts of the action leaming process at the expense of limiting
other parts.
“Research on leaming styles has shown that managers on the whole are
distinguished by very strong active experimentation skills and are very weak
on reflective observation skills” (Kolb, Rubin and Osland, 1991: 67).
The difficulty inherent in managing amidst high levels of complexity has been described
by Doemer (1980), while on the other hand, cautionary wamings of the effects of
misperceptions of dynamic feedback, and information processing limitations have been
issued by Sterman (1994) -- though some choose to ignore these signs altogether. Even
recommendations that policy makers should decide whether they wish to make decisions
based on results obtained from using their unaided mental models, or altematively, from
results obtained from some combination of formal and mental models” (Radzicki 1998
:8) are not heeded by practitioners.
An open systems view of organization is related to the interaction of the organization and
its environment. As such, organizations have two sets of complexity to deal with, those
that represent intemal dynamics and those that deal with extemal environmental changes.
The ability to deal with complexity requires the ability to leam from observed pattems of
change over time. In tum, groups that leam and engage in collective sense-making
together should, by all standard accounts, be more effective. That is, more able to
improve effectiveness of their decisions in such complex environments by leveraging the
value of contemporary tools, techniques, and technologies to their advantage. For
example, developing shared mental models, leaming as a group, and using systems
thinking comprise some of the core elements suggested by Senge (1990) as being
necessary for performance improvement and in leaming organizations. Can the same
sorts of tools be applied with equal benefit by leamers in the process of developing
greater capabilities for systems thinking? Is there a way of teaching people to become
better systems thinkers that accounts for individual leaming needs and preferences?
Integrating Tools for Modeling and Learning
It is common among system dynamics practitioners to assume that all systems thinking
tools have equal values to leamers. However, leamers may have preferences to choose
certain tools over others due to their intrinsic leaming styles. If the assumption of all
systems tools having equal values to leamers is unjustified, it portends the need for a type
of meta-model for leaming systems thinking that can acdress variations in leaming style.
Part of the process of system thinking and systems dynamics requires the
conceptualization of a working hypothesis for the system under study and the
development of a reference mode for a simulation model. For example, Saeed (2003) has
suggested that “the conceptualization of a reference mode requires the same leaming
process as the development of a dynamic hypothesis, the construction of a model, the
creation of the model understanding and the design of a policy for system improvement”.
In this paper, Saeed reverses the directions originally envisioned by Kolb and creates a
means of connecting thought and action. Using the Kolb model it was suggested that
careful observation of facts and pattems should drive thinking. That the interpretation of
these facts must be tested through experimentation and that the results be tumed into
generalizations that could then be reused to start a new leaming cycle. (Saeed, 2003).
Figure 1 is extracted from Saeed’s diagram, to which was added the Leaming Style
associated with the Kolb quadrants of orientation.
Systems
Boundary Concrete
Experience
Accommodator concrete Diverger
active passive
Graohi Historical
raphing Data
Pattems
Reflective
. Assimilator Observation
Active Converger
Experimentation. " 7
NO Reference
Mode
Abstract
Conceptualization
Fig 1 Learning Styles and Model Development Process
(From Saaed 2003 and Kolb, 1991)
(Note: The Kolb model nonmally has the direction of arrows moving clockwise).
Lawrence (1995) uses the concept of leaming style to describe four aspects of
psychological makeup. These are:
e A preferred or habitual pattem of mental functioning, information processing,
formation of ideas and judgments.
e Pattems of attitudes and interests that influence what a person will attend to ina
potential leaming situation.
e A disposition to seek out leaming environments compatible with one’s cognitive
style, attitude and interests.
e A disposition to use certain leaming tools, to use them successfully and to avoid
other tools. (:39).
It is this final characteristic that serves as a key block of the foundation for the research
conducted for this paper. It is proposed that people exhibit pattems of leaming from
experience that reflect the manifestation of certain proclivities to seek information, assign
meaning, and attend to certain cues at the exclusion of others. These tendencies often
minor personal beliefs and intrinsic traits that differentiate people, yet whose origins are
not well understand. Koestler (1945), for example, identifies leaders as holding
worldviews that tend toward being belonging to a constellation of attributes which he
describes as being either akin to that of a Yogi ora Commissar. Ina similar way, other
typologies distinguish between natural human tendencies to perceive the world, make
sense, and leaming in quite divergent ways that we often refer to as leaming styles.
Learning Styles
In order to evaluate leaming style, several inventories are available including those
developed by Honey and Mumford (1995), Dechant and Marsick (1993), Murrell (1987)
and Glacel (2001). However, the inventory selected for use in this study was Kolb’s
KLSI (1991) which chosen based on the work Saeed (2003). The interpretation of scores
is given as an orientation while combined orientations give a resulting leaming style. The
following is a summarization of the orientations developed from the Kolb leaming style
inventory model.
Orientation _I.
An orientation toward concrete experience focuses on being involved in experiences and
dealing with immediate human situations. There is an emphasis on feelings and doing, or
being in an experience.
Orientation II
An orientation towards reflective observation focuses on understanding the meaning of
ideas and situation. It emphasizes understanding as opposed to practical application and
thinking over doing.
Orientation III
An orientation towards abstract conceptualization focuses on using logic, ideas and
concepts. Persons with this orientation are good at systematic planning.
Orientation IV
An orientation to toward active experimentation focuses on actively influencing people
and changing a situation. There is an emphasis on practical application. There is a value
in the ability to impact the environment. (Kolb, et. al., 1991: 62) A person s leaming style
is based on the balances achieved by the orientations.
Systems Thinking and Learning Styles
When the potential impact of leaming style on systems thinking is considered, a number
of different possibilities emerge for pattems of capacity development.
Convergent Style
Convergent styles enjoy problem solving and prefer technical tasks and seem to do best
where there is a single correct answer or solution (Kolb, 1991). It is proposed here that
the problem with such styles in systems thinking is that there can be several solutions to a
systems problem, depending on the leverage points that are identified. Could this style
have difficulty in using any of the tools or in the development of system thinking skills?
Divergent Style
The divergent style emphasizes concrete experience and reflective observation. There
strength lies in awareness of meaning and values. The primary adaptive ability on this
style is to view concrete situations form many perspectives and to organize many
relationships into a meaningful “Gestalt” (sic: whole). One might suspect that the
attraction to such a style would be the initial development of behavior over time graphs
as a way of viewing the situation.
Assimilation Style
Assimilation is a style where abstract conceptualization and reflective observation are
preferred. The strength of this style lies in deductive reasoning and the ability to create
theoretical models. Could this style be more attracted to the development and use of
causal loop diagramming or the process of connecting the dots, so to speak, in the picture
of the system?
Accommodation Style
Finally, the accommodation style has emphasizes concrete experience and active
experimentation. The strength lies in doing things. The emphasis is on risk taking. Does
this style enjoy playing the game and using direct feedback? (Kolb, et. al., 1991)
Research
An experiment was conducted in order to determine the relationship that may exist
between action leaming styles and the choice of system thinking tools. Data was
collected from a sample using a questionnaire designed to assess individual preferences
for systems thinking tools and leaming styles and types. The sample was taken from
undergraduate management students at a university.
Sample and Treatment
The sample was composed of fifty students who were registered in a course on
systems thinking for managers. These students volunteered to participate in this research
and understood that their participation in the study would not influence their course
grade. Prior to completing the questionnaire used for this research students were
provided with approximately fifteen hours of training in the use of the ten systems
thinking tools. They also read the first seven chapters of the book by Peter Senge (1990)
The Fifth Discipline. Finally, they played several management flight simulators,
including Thompson’ s (2004) Two Brothers Pizza and Sterman’s (199X) B&B
Enterprises individual and as part of a team for a period of approximately nine hours
each.
Students were asked in the questionnaire to rate the frequency of use and
perceived value of each of the ten systems thinking tools in contributing to profitability in
the B&B Enterprises flight simulator. The rating scale used was not a forced choice scale
and did not require that student choose one tool as being more important than any others.
It is acknowledged that systems thinking tools are usually used in conjunction with each
other and this will potentially produced interactive effects. However, the purpose of this
research is to simply explore the following questions: 1. Are there preferences in the
choice and use of some systems thinking tools, and 2. Are preferences for systems
thinking tools influenced by individual leaming styles, and 3. Is performance, as
measured by profit, ina management flight simulator influenced by more frequent use of
certain systems thinking tools.
Method
In order to evaluate if a relationship exists between leaming styles and a “disposition to
use certain systems thinking tools, to use them successfully and to avoid other tools.
(:39)”, research was conducted using the Kolb Leaming Style Inventory (KLSI). The
same students had completed approximately nine hours of playing just the B& B
Enterprises Microworld -- developed by Sterman (www.forio.com). The questionnaire
was designed to capture the types of system thinking tools available for use in the game,
and the frequency of use of these systems thinking tools. It is hypothesized that
frequency of use is one indicator of system tool preference - thought not exclusively so.
Frequency of use may indicate preference, but there are other factors that contribute to
preference, such as tool simplicity, ease of use, and perceived value relative to other
tools. Also, by their intrinsic nature some tools, such as assigning link polarities, are
generally used more often than tools, such as drawing Behavior Over Time graphs.
However, it is also equally feasible that tools with relative advantages in terms of ease of
use or perceived value are likely to be used more often. Since this is an exploratory
student in uncharted waters, our initial effort was streamlined in order to focus on the
defining the basic relationships in the study. Finally, the questionnaire also measured of
economic performance (cumulative profit) in the B&B flight simulator was recorded for
both individuals and teams.
Both sets of data, from the KLSI and the records of decision made to select certain
system thinking tools were analyzed. Participants were grouped in several ways. First,
participants were grouped according to their leaming styles, (i.e. Accomodator,
Converger, Diverger, or Assimilator). A second classification was created using the
orientations that make up each style, such as whether a participant prefers to act by using
concrete experience or active experimentation, or prefers abstract conceptualization to
engaging in reflective observation. The results of the analysis are discussed below.
Results
Introduction
Being an exploratory research investigation, the data were analyzed in a logical process
to assess the hypotheses and research questions. Initially, the frequency and preference
of use of the various tools, across the various leaming styles were analyzed. Due to the
small sample sizes in several leaming style categories, the additional analyses were
conducted used Kolb’s dichotomies of concrete vs. conceptual and active vs. passive
categories. For each of these dichotomies the frequency of use and preference of use
were analyzed.
Part I
The first analysis looked at the rating of the frequency of use of the various tools across
all leaming styles. The results are shown in Table 1, which shows the average (mean)
and standard deviation for each style. The Analysis of Variance indicates an F =8.262
(df =9, 420), which is significant beyond the .01 level. Thus the data indicate that there
are differences amongst the various tools based on the rating of the frequency of use.
SUMMARY Frequency of Use — All styles
Standard.
Groups Count Sum Average Deviation
FCausLoop 43 165 3.837 1.194
FTimeDelay 43 «173 4.023 0.886
FBOT 43 142 3.302 1.166
FSysArch 43 138 3.209 1.125
FPolarities 43 155 3.605 1.116
FCausEffec 43 185 4.302 0.741
FStckHlow 43° «116 2.698 1.186
FMental 43 157 3.651 1.131
FDiscov 43 «171 3.977 0.988
FKeepLog 43. 162 3.767 0.972
ANOVA
Source of Pp-
Variation ss df MS F value Fcrit
Between
Groups 83.54419 9 9,283 8.262 0.000 1.902
Within Groups 471.8605 420 1.123 Signif
Total 555.4047 429
Part II
The next analysis looked at the rating of the preference of the various models across all
leaming styles. The results are shown in Table 2, which shows the average (mean) and.
standard deviation for each style. The Analysis of Variance indicates an F = 6.689 (df =
9, 420), which is significant beyond the .01 level. Thus the data indicate that there are
differences amongst the various tools based on the rating of the preference of the tools.
Table 2
SUMMARY Preference of Method — All styles
Standard
Groups Count Sum Average Deviation
PCausLoop 433 175 4.070 1121
PTimeDelay 43 181 4.209 0.773
PBOT 43 164 3.814 0.958
PSysArch 43 147 3.419 1.118
PPolarities 43 163 3.791 1.125
PCauseEffect 43 190 4.419 0.731
PStckFlow 43 130 3.023 1.080
PMental 43 162 3.767 1.065
PDiscov 43 173 4.023 0.938
PKeepLog 43 166 3.860 1.087
ANOVA
Source of
Variation ss df MS F P-value Farit
Between Groups 60.67209 9 6.741 6.689 0.000 1.902
Within Groups 423.2558 420 1.008 Signif
Total 483.9279 429
Due to very small sample sizes in many categories no analyses were performed for
frequency of use or preference of methods for individual leaming styles. Instead the
following sets of tables report the analyses based on the dichotomy of categories of
conceptual vs. concrete and active vs. passive.
Part IIL.
Table 3 presents the analysis for the rating of frequency of use for a conceptual
orientation. The data in Table 3 shows the average (mean) and standard deviation for
each tool. The Analysis of Variance indicates an F = 1.429 (df =9, 70), which is not
significant beyond the .01 level. The data indicate that there are no differences amongst
the various tools based on the rating of the frequency of use of the various tools.
Part IV
Table 4 present the analysis for the rating of frequency of use for a concrete orientation.
The data in Table 4 shows the average (mean) and standard deviation for each tool. The
Analysis of Variance indicates an F =7.628 (df =9, 340), which is significant beyond the
.01 level. Thus the data indicate that there are differences amongst the various tools
based on the rating of the frequency of use of the various tools for those individuals who
have a concrete orientation.
Table 3
Frequency of
SUMMARY Use - Conceptual
Standard
Groups Count Sum _ Average Deviation
FCausLoop 8 27 3.375 1.598
FTimeDelay 8 30 3.750 1.085
FBOT 8 20 2.500 1.512
FSysArch 8 2 3.250 1.389
FPolarities 8 30 3.750 1.282
FCausEffec 8 32 4.000 0.756
FStckHlow 8 20 2.500 1.414
FMental 8 2 3.625 1.302
FDiscov 8 30 3.750 1.282
FKeepLog 8 31 3.875 0.991
ANOVA
Source of P-
Variation ss df MS F value Frit
Between
Groups 21.0625 9 2.340 1429 0.193 2.017
WithinGroups 114625 70 1.638 Non-Signif.
Total 135.6875 79
SUMMARY Frequency of Use - Concrete
Standard.
Groups Count Sum _ Average Deviation
FCausLoop 35 (138 3.943 1.083
FTimeDelay 3 14 4.086 0.853
FBOT 35 «122 3.486 1.011
FSysArch 35 «112 3.200 1.079
FPolarities 35 «(125 3.571 1.092
FCausEffec 35 (153 4371 0.731
FStckHlow 35 «(96 2.743 1.146
FMental 35 (128 3.657 1.110
FDiscov 35 «141 4.029 0.923
FKeepLog 35. 131 3.743 0.980
ANOVA
Source of Pp-
Variation ss df MS F value Farit
Between
Groups 69.85429 9 7.762 7.628 0.000 1.907
Within Groups 345.9429 340 1.017 Signif
Total 415.7971 349
Part V
Table 5 present the analysis for the rating of preference for tools for those who have a
conceptual orientation. The data in Table 5 shows the average (mean) and standard
deviation for each tool. The Analysis of Variance indicates an F = 1.675 (df =9, 70),
which is not significant beyond the .01 level. Thus, the data indicate that there are no
differences amongst the preferences among the various tools for those with a conceptual
orientation.
Part VI
Table 6 present the analysis for the rating of preference tools for those with a concrete
orientation.. The data in Table 4 shows the average (mean) and standard deviation for
each tool. The Analysis of Variance indicates an F =6.009 (df =9, 340), which is
significant beyond the .01 level. Thus the data indicate that there are differences amongst
the preferences among the various tools for those individuals who have a concrete
orientation.
Table5
SUMMARY Preference of Method - Conceptual
Standard
Groups Count Sum Average Deviation
PCausLoop 8 30 3.750 1.581
PTimeDelay 8 29 3,625 0.916
PBOT 8 23 2.875 0.835
8 26 3.250 1.282
PPolarities 8 30 3.750 1.488
PCauseEffect 8 XA 4250 1.165
PStckFlow 8 19 2375 0.916
PMental 8 29 3,625 1.188
PDiscov 8 29 3,625 1.188
PKeepLog 8 32 4000 1.195
ANOVA
Source of
Variation ss df MS F P-value Fait
Between
Groups 21.6125 9 2401 1.675 0.112 2.017
Within Groups 100.375 70 1.434 Non-Signif.
Total 121.9875 79
SUMMARY Preference - Concrete
Standard.
Groups Count Sum Average Deviation
PCausLoop 35.145 «844143 1.004
PTimeDelay 35.152 4.343 0.684.
PBOT 35 «141 4.029 0.857
35 (121 3.457 1.094
PPolarities 35 133 3.800 1.052
PCauseEffect 35 156 4.457 0.611
PStckFlow 35 (111 3.171 1.071
PMental 35 133 3.800 1.052
PDiscov 35 144 4114 0.867
PKeepLog 35.134 3.829 1.014
ANOVA
Source of
Variation ss df MS F P-value Fait
Between
Groups 48,228571 9 5.359 6.009 0.000 1.907
Within Groups 303.2 340 0892 Signif
Total 351.42857 349
Part VIL.
Table 7 present the analysis for the rating of frequency of use for the active orientation.
The data in Table 7 shows the average (mean) and standard deviation for each tool. The
Analysis of Variance indicates an F =3.002 (df =9, 140), which is significant beyond the
.01 level. Thus the data indicate there are differences amongst the various tools based on
the rating of the frequency of use of the various tools within an active orientation.
Part VIII.
Table 8 present the analysis for the rating of frequency of use for the passive orientation.
The data in Table 8 shows the average (mean) and standard deviation for each tool. The
Analysis of Variance indicates an F =5.380 (df =9, 297), which is significant beyond the
.01 level. Thus the data indicate that there are differences amongst the various tools
based on the rating of the frequency of use of the various tools for those individuals who
have a passive orientation.
Table 7
SUMMARY Frequency of Use - Active
Standard
Groups Count Sum Average Deviation
FCausLoop 15 58 3.867 1.356
FTimeDelay 15 @ 4.133 0.640
FBOT 15 2 3.467 1.302
FSysArch 15 49 3.267 1.223
FPolarities 15 57 3.800 1.014
FCausEffec 1 6&8 4.200 0.941
FStckFlow 15 39 2.600 1.352
FMental 15 56 3.733 1.033
FDiscov 15 @ 4.133 0.915
FKeepLog 15 60 4.000 1.069
ANOVA
Source of
Variation ss df MS F P-value Fart
Between
Groups 33.04 9 3.671 3.002 0.003 1947
Within Groups 171.2 140 1.223 Signif
Total 204.24 149
Table
Frequency of Use -
SUMMARY Passive
Standard
Groups Count Sum Average Deviation
FCausLoop 28 107 3.821 1.124
FTimeDelay 28 111 3.964 0.999
FBOT 28 90 3.214 1.101
28 «89 3.179 1.090
FPolarities 28 «98 3.500 1.171
FCausEffec 28 122 4.357 0.621
FStckFlow 28 «#77 2.750 1.110
FMental 28 101 3.607 1.197
FDiscov 28 109 3.893 1.081
FKeepLog 28 102 3.643 0.911
ANOVA
Source of Pp-
Variation ss df MS F value Fart
Between
Groups 59.96104 9 5.996 5.380 0.000 1.863
Within Groups 331.0357 297 1.115 Signif
Total 390.9968 307
Part IX
Table 9 present the analysis for the rating of preference of tools for an active orientation.
The data in Table 9 shows the average (mean) and standard deviation for each tool. The
Analysis of Variance indicates an F =3.226 (df =9, 140), which is significant beyond the
.01 level. Thus, the data indicate that there are differences amongst the various tools
based on the rating of the preference amongst the various tools for those with an active
orientation.
Part X
Table 10 present the analysis for the rating of preference of the method for those with a
passive orientation. The data in Table 10 shows the average (mean) and standard
deviation for each tool. The Analysis of Variance indicates an F =3.886 (df =9, 270),
which is significant beyond the .01 level. Thus the data indicate that there are differences
amongst the various tools based on the rating of the preferences amongst the various tools
for those individuals who have a passive orientation.
Table 9
Preference of use -
SUMMARY Active
Standard
Groups Count Sum Average Deviation
PCausLoop 15.000 60.000 4.000 1.134
PTimeDelay 15.000 62.000 4.133 0.640
PBOT 15.000 59,000 3.933 0.961
15.000 51.000 3.400 1.298
PPolarities 15.000 59,000 3.933 1.100
PCauseEffect 15.000 70.000 4.667 0.617
PStckFlow 15.000 45.000 3.000 1.069
PMental 15.000 60.000 4.000 1.069
PDiscov 15.000 65.000 4.333 0.724
ANOVA
Source of
Variation ss df MS F P-value Frit
Between
Groups 28.593 9.000 3177 3.226 0.001 1.947
Within Groups 137.867 140.000 0.985 Sig.
Total 166.460 149,000
Table 10
SUMMARY Preference of Use - Passive
Standard.
Groups Count Sum Average Deviation
PCausLoop 28.000 115.000 4107 1.133
PTimeDelay 28.000 119.000 4250 0.844
PBOT 28.000 105.000 3.750 0.967
28.000 96.000 3429 1.034
PPolarities 28.000 104.000 3.714 1.150
PCauseEffect 28.000 120.000 4286 0.763
PStckFlow 28.000 85.000 3.036 1.105
PMental 28.000 102.000 3.643 1.062
PDiscov 28.000 108.000 3.857 1.008
PKeepLog 28.000 106.000 3.786 1.031
ANOVA
Source of
Variation ss df MS F P-value Frit
Between
Groups 36.143 9000 4016 3.886 0.000 1915
Within Groups 279.000 270.000 1.033 Sig.
Total 315.143 279.000
Analysis
This preliminary evaluation revealed that there was no apparent relationship between
action leaming styles, as defined by the KLSI, such as Diverger, Acccomodator,
Converger or Assimilator and preference for certain systems thinking tools. To confirm
that no relationships might have been missed, the participant groups were disaggregated.
into the four major orientations such as; “concrete-experience”, “active-experimenter,”
“reflective- observer”, and “abstract -conceptualizer’. A second statistical analysis was
then carried out for these groupings. The subsequent evaluation of this data revealed that
the following findings:
1. There is evidence of significant differences in the rating of the frequency and
preferences of use of a given tool amongst an individual’ s orientation within the Kolb
model.
2. Within orientation groupings, there are significant relationships among the group
orientation and both preference and frequency of use of systems thinking tools
3. Differences within groups existed for all groups except the abstract conceptualizers.
Discussion.
The evaluation of action leaming styles indicated differences among the Kolb Leaming
orientations and the systems thinking tools of choice. Before the exact nature of this
relationship can be discemed, it is recommended that further research be conducted in
this area. In addition, the nature of any relationships should be analyzed in order to
determine the relationship between stylistic groupings. Further, an analysis of the
relationship between frequency of use and preferences for the various tools is warranted.
The study revealed significant differences in preference for use of the ten various systems
thinking tools among participants. However, there was no significant relationship found
between use of preferred systems thinking tools and flight simulator performance. (as
measured by cumulative profit)
If an analysis reveals a relationship between group orientations and systems thinking
tools, the final results could lead to some more solid conclusions that might suggest a
more definitive relationship between leaming orientations and preference for using
certain systems thinking tools over others. The existence of such a well-defined
relationship would be significant to the practice of system dynamics because it would
enable instructors or course designers to develop methods capable of “hooking”, the
understandings of the various groups by aligning tool use with their natural action
leaming orientation. This would be a similar approach to the use of language as a means
of interacting with Myers-Briggs types. Groups could also be clustered according to
systems thinking tool preferences to simplify or expedite systems thinking instruction.
“Buckley and Caple (cited in Grace2007) have described the leaming styles in their
book The Theory and Practice of Training and mace some observations on the validity
of both models because of their reliance on the term ‘experience’ and the fact we lack a
Clear understanding of what an ‘experience’ is. From the viewpoint of training the real
relevance must surely be that different people do seem to leam more effectively in
different ways. For some the concept of leaming by activity (activist) is by far the most
attractive, while others are much more reflective (reflectors) by nature and prefer to be
given information and allowed time to think about it before doing anything” (Grace,
2007). In his paper Grace goes on to describe both favorable and unfavorable activities
for each leaming style.
Finally, facilitators could choose which tools to select for presenting the ideas of system
dynamics to groups. The core of the proposed concept is that by selecting systems
thinking tools that homogeneous groups both prefer and use freely, participants would be
more likely to accept the ideas of system thinking and system dynamics as tools for use in
problem applications.
References
1. Cavaleni, S. (2002). “Evaluating the Performance Efficacy of Systems Thinking Tools”,
Proceedings of the 20" Intemational Conference of the Systems Dynamics Society. 2002.
Palermo, Italy. 57.
2. Cavaleri, S. and Sterman, J. (1997). “Towards evaluation of systems thinking interventions:
acase study”, Systems Dynamics Review. Vol.13. No.2, Summer 171-186.
3. Cavaleni, S and Friedman, S. (2003). “Evaluating Changes in System Thinking
Capacity: A Methodology Based on Alpha, Beta, Gamma Analysis”, Proceedings of
the 21° International Conference of the Systems Dynamics Society, New Y ork, New
York.
4, Cavaleri, S. and Seivert, S. (2005). Knowledge Leadership, Boston, Butterworth-
Heinemann.
5. Dechant, K. and Marsick, V. (1993). “The Team Leaming Survey”. King of Prussia,
PA. The HRD Quarterly.
6. Doemer, D. (1980). “On the Difficulty people have in Dealing with complexity”.
Simulation and Games. Vol.11. No.1. March.
7. Doyle, J. (1997). “The Cognitive Psychology of Systems Thinking”, System Dynamics
Review. Vol. 13. No. 3, Fall, 253-265.
8. Durham, J.L.R. (2002). “Balanced Scorecards, Mental Models and Organizational
Performance. A Simulation Experiment”, Ph.D. Dissertation, University of Texas at Austin.
9. Durham, J.LR. (2003). “The Systematic Influence of Scorecards on Mental Models and.
Organizational Performance”, Proceedings of the System Dynamics Winter Camp. 2003. The
University of Texas. Austin Texas.
10. Edwards,W and Winterfeldt.D.Von. (1986). “On cognitive illusions and their
implications.” (In Judgment and Decision Making an Interdisciplinary Reader. Ed.
H. Arkes and K. Hammond, New Y ork, Cambridge University Press).
11. Glacel, B.P. (2001).” A Survey forTeam Leaming’. The 2001 Annual: Volume 2.
Consulting, Jossey Bass/Pfeiffer. P.117
12. Grace, M. Leaming styles. British Dental Joumal. London, England.
hittp://www.nature.conybdi/joumal/v191/n3/full/4801116a. html. Retrieved May 22,
2007. (Note: Caple, J. and Buckley, R. (2007). The Theory and Practice of Training.
New York. Kogan Pub.)
13. Graham A. and Senge, P. (1990). “Computer Based Case Study and Leaming
Laboratory Projects”, System Dynamics Review. Vol. 6, Winter, 100-105.
14. Honey , P. and Mumford, A. (1995). “Capitalizing on Y our Leaming Style”. King of
Prussia, PA. Organizational Design and Development.
15. Huz, S, Anderson, D. F., Richardson, G.P. and Boothroyd, R. (1997). “A Framework for
evaluating systems thinking interventions: an experimental approach to mental health systems
change”, Systems Dynamics Review, Vol. 13. No.2, Summer, 149-169.
16. Kim, D. H. (1994). Systems Thinking Tools, Cambridge, MA, Pegasus Communications.
17. Koestler, A. (1945). The Yogi and the Commissar, London, Jonathan Cape Ltd.
18. Kolb,D.A., Rubin, I.M. and Osland, J. (1991). Organizational Behavior. Englewood
Cliffs, NJ. Prentice Hall
19. Lawrence, G. (1995). “People types and tiger stripes”. Gainesville Horida.Center for
Psychological Type.
20. Morecroft, J. and Sterman. J.(1994), Modeling for Learning Organizations,
Productivity Press.Portland Oregon. ( Currently Pegasus Communications).
21. Mune, K.L. (1987). “The Leaming Model Instrument: An Instrument Based on The Leaming
Model for Managers”. The 1987 Annual: Developing Human Resources. UA Associates.
22. Pugh, A. and Richardson, G. (1981). Introduction to Systems Dynamics Modeling
with Dynamp, Cambridge, Massachusetts. Productivity Press.
23. Radizicki, M. (1998). Introduction to System Dynamics Modeling (Draft). Worcester
Polytechnic Institute.
24. Richmond, B. (1993). “Systems thinking; critical thinking skills for the
1990s and beyond”, System Dynamic Review, Volume 9 Number 2 Summer. 113-
133.
25. Saeed, K. (2003). Articulating developmental problems for policy intervention: A
system dynamics modeling approach. Simulation and Gaming. 34(3): 409-436
26. Senge, P. (1990). The Fifth Discipline, New Y ork, Doubleday/Cumrency.
27. Sterman, J. (1994). “Leaming in and About Complex Systems”. MIT Paper D-4428.
Cambridge, Massachusetts, Massachusetts Institute of Technology.
