A Model of Educational Innovation
Gary B. Hirsch
Consultant and Creator of Learning Environments
Wayland, Mass.
The Challenge of Innovation in Schools
It is difficult to implement innovations in schools. Everyone favors innovation, but focuses on
different things depending on what they believe is the “magic bullet” for improving schools.
Discussions of innovation are often confused because people lump together many different types
of innovations including new curricula and methods of teaching, ways of evaluating students, and
methods of managing schools. Even though a number of innovations are being considered or
implemented at the same time, they are often treated in isolation from each other as well as from
what is already going on in the school. Different kinds of innovations compete with each other
for resources with the frequent result that none really gets the resources and attention it needs to
be successfully implemented. Rather than replacing the existing curricula and systems,
innovations are often added to what is already in place. As a result, innovations in schools often
create “pushback” in the form of resistance from overworked teachers and students. Teachers
may feel that innovations are attempts to “fix something that is not broken” and represent change
for change’s sake. Their fears are often confirmed when widely-touted innovations are replaced
by the next fad a few years later.
Background of This Effort
How can innovations be planned and implemented to avoid these problems and produce an
effective impact? This paper describes a model of innovation in schools developed to provide a
framework for discussing and understanding these problems. The model has also been used to
answer some “What if?” questions about different approaches to implementing innovations and
produce some simple rules of thumb.
The model grew out of conversations between Jay Forrester and Ted Sizer who leads a group
called the Coalition of Essential Schools and has written extensively on change in schools. Gary
Hirsch got involved and, under a grant from the Gordon Stanley Brown Fund, set out to model
innovation in schools. A group that included Forrester and Sizer and a number of other
educators guided the work and identified critical factors affecting innovation in schools.
Additional insights were provided by a small group of educators actively involved in applying
System Dynamics in their schools.
The work is in an early stage. The model is presented to elicit reactions from the System
Dynamics community and from educators and especially from people who are in both camps.
This paper presents some of the model’s structure, results of simulations with the model, and
implications for implementing innovation in schools drawn from those simulation results. (A
more complete description of the model and simulation results entitled “Innovation in Schools: A
Model to Help Structure the Discussion and Guide the Search for Strategies” (D-4765) is
available from the Creative Learning Exchange at http://sysdyn.mit.edu/cle.)
Modeling Innovation in Schools
1. Curriculum Innovation Process
One part of the model focuses on the process of innovation itself. As shown in Figure 1, the
model depicts innovation in schools as a multi-stage process with several factors at each stage
that affect the extent to which a curriculum innovation is adopted and has impact. (In this
discussion, model variables appear in Title Case.) Users of the model can specify the magnitude
of an innovation as a function of its Breadth of Applicability (whether it applies to a small group,
an entire grade, or an entire school system), Rote Memorization vs Deeper Understanding, and
Generality of Applicability (to a specific subject area or a set of skills applicable to a broad range
of subjects). Resources required for implementing an innovation will depend on its magnitude as
well as the Distance from Current Practice that the innovation represents.
Level of Professional Development Mode of Student Evaluation
Flexibility Given Teachers Re Innovation \
Measured Impact of Innovation
Ability to Tailor Innovations to Students Learning Styles
Teacher Motivation
Structural Flexibilit
Teacher Initiated Innovations
_ Adopted
Innovations in Process Curriculum snd tan Adopted — Impact of Innovation
Systemwide aN So
” Students Capacity for Learning
Hours Required for Curriculum Innovatio!
Hours Available for Curriculum Innovation
Figure 1: Curriculum Innovation Process
As shown in Figure 1, new curricula may be introduced as Systemwide Innovations and/or
Teacher Initiated Innovations. The rate of Teacher Initiated Innovations depends on a climate
affected by both Teacher Motivation and the Flexibility Given Teachers Re: Innovations. New
curricula introduced by either route become Innovations in Process and can eventually, if
conditions are right, become Curriculum Innovations Adopted. The Fraction Adopted (of
Innovations in Process) is also a function of Teacher Motivation as well as the Structural
Flexibility ina school. Structural Flexibility might, for example, he reflected by a scheduling
system that includes different-sized blocks of time rather than a rigid, single-sized block that is
inadequate for more elaborate student- directed, leaning experiences. Low Teacher Motivation
and/or low Structural Flexibility can cause most of an innovative curriculum to be discarded
rather than being adopted. The Rate of Adoption is also affected by the ratio of Hours Available
for Curriculum Innovation to the Hours Required for Curriculum Innovation. Having more hours
available will permit a faster Rate of Adoption.
Adoption of curriculum innovations does not guarantee that they will have an impact. The
Impact of Innovation will depend on both the level of Curriculum Innovations A dopted and the
Ability to Tailor innovations to Students Learning Styles. The Ability to Tailor Innovations to
Students Learning Styles depends, in tum, on Flexibility given Teachers Re: Innovation and
teachers’ Level of Professional Development. These factors suggest that innovations cannot
simply be implemented in lockstep fashion and be expected to have an impact on students.
Teachers need the flexibility to tailor innovations and the skill to adjust curricula to meet student
needs.
The Impact of Innovation has its effect in two places. One is directly on Students Capacity for
Learning and the other is on the perception of teachers and the community of the impact,
expressed as the Measured Impact of Innovation. Why differentiate between the Impact of
Innovation and Measured Impact of Innovation? Many curriculum innovations will produce
impacts that are largely invisible because the Mode of Student Evaluation in use does not reveal
the impact. For example, an innovation may produce an increase in higher-level skills. If the
Mode of Student Evaluation in effect is merely traditional testing of rote memorization, these
higher-level skills will be invisible and, in this case, the Measured Impact of Innovation will be
negligible even though the Impact of Innovation is significant.
2. Community and Teacher Responses to Innovation
Trust Between School and Community
Performance on Traditional Curriculum
Patience with Innovation
Level of Expected Impact—a Experience with Innovation Impact of Innovation
Conflict with Teachers
Measured Impact of Innovation
Innovations Initiatet
\ Mode of Student Evaluation
Teacher Motivation
AN *___ Peroeived Need for Innovation
Stress on Teachers ~~
Awareness of Innovation Elsewhere Students Capacity for Learning
Level of Professional Development
Figure 2: Community and Teacher Response to Innovation
The other part of the model deals with how teachers and other members of the community
respond to innovation and how this, in turn, affects the future climate for and acceptance of
innovation. Teacher Motivation is critical for assuring that any innovation will be adopted. Low
motivation will mean that teachers will only “go through the motions” in working with an
innovative curriculum and an innovation will have minimal impact on how students leam. The
group identified several factors that can affect Teacher Motivation. Experience with Innovation
is especially important. As shown in Figure 2, Experience with Innovation changes depending
on how the Measured Impact of Innovation compares with the Level of Expected Impact.
Perceived Need for Innovation is another factor affecting motivation. Teacher Motivation will be
low unless teachers see an innovation as meeting a real need. Perceived Need for Innovation also
depends on the Mode of Student Evaluation since a need may exist, but be invisible if the
measurement system in place does not reveal the need. Again, a Mode of Student Evaluation
focused on rote memorization will not reveal inadequacies in higher level skills. Teacher
Motivation may also be greater if teachers have an A wareness of Innovation Elsewhere that can
come from a higher Level of Professional Development. Teacher Motivation can also be
undermined by Stress on Teachers if they are forced to spend extensive numbers of hours
adopting innovative curricula in addition to their ongoing duties.
The community’ s response is as important as that of teachers. The group identified Trust
Between School and Community as a key variable affecting innovation. Experience with
Innovation has as an important effect on this variable as it does on Teacher Motivation. Conflict
with Teachers arising from Stress on Teachers will also undermine Trust Between School and
Community. Performance on Traditional Curriculum will also affect Trust Between School and
Community. Diverting too much time from the traditional curriculum to work on innovative
curricula will provoke a negative reaction from the community if Performance on Traditional
Curriculum falls before the benefits of innovation can be realized. A higher level of Trust
Between School and Community will increase the community’ s Patience with Innovation which
will enable the Level of Expected Impact to develop more slowly and allow more time for the
Measured Impact of Innovation to grow and favorably affect Experience with Innovation.
3. Overview: Putting Innovation and Teacher and Community Response Together
Figure 3 puts key aspects of the innovation process and responses to innovation together to
provide an overview of the complete model. It shows how Trust Between School and
Community can have an important effect on changing the conditions that affect innovation. A
high level of Trust Between School and Community is assumed to be necessary for making any
changes in the Mode of Student Evaluation and in Structural Flexibility. In the absence of
sufficient trust, communities are likely to react conservatively to proposed changes and want to
stick with a “tried and true” approach to running a school. A high level of Trust Between School
and Community may also make it possible to cut back on Tasks in Traditional Curriculum and
Hours Required for Traditional Curriculum and leave more Hours Available for Curriculum
Innovation.
Trust Between School and Community
Level of Expected Impact { Mode of “y Evaluation
Experience x Innovation
Measured so of Innovation
~
Tasks in Traditional Curriculum
Students Capacity for Learning
Impact of Innovation
\ Motivation
Curriculum Innovations Initiated
Structural Flexibility
Curriculum Innovations in Process
a wa
Ss oa Tnnovations A dopted
Hours Required for Traditional Curriculum
Hours Available for Curriculum Innovation
Figure 3: Overview: Putting Community and Teacher Response Together
The relationships shown in Figure 3 contain several reinforcing loops that, if hamessed
appropriately, can help schools overcome the inertia that usually serves as an impediment to
innovation. In one set of loops, positive Experience with Innovation will lead to higher levels of
Trust Between School and Community that make it possible to change the Mode of Evaluation
and increase Structural Flexibility. These changes can then lead to a higher Fraction A dopted
and Measured Impact of Innovation which produce further improvements in Experience with
Innovation. Being able to cut back on Tasks in Traditional Curriculum as a result of higher Trust
Between School and Community will reduce Hours Required for Traditional Curriculum and
make more Hours Available for Curriculum Innovation. This will, in tum, lead to more
Curriculum Innovations A dopted, greater Impact of Innovation and Measured Impact of
Innovation, and further improvement in Experience with Innovation and Trust Between School
and Community.
Improvements in Students Capacity for Learning will also reduce Hours Required for Traditional
Curriculum (i.e., students will be able to leam more efficiently), increase the Hours Available for
Curriculum Innovation, Curriculum Innovations A dopted, and Impact of Innovation and further
increase Students Capacity for Learning. Increases in Students Capacity for Learning can also
increase the Impact of Innovation which can then produce further increases in Students Capacity
for Learning. Finally, positive Experience with Innovation will increase Teacher Motivation
which will, in turn, lead to greater Curriculum Innovations A dopted, Impact of Innovation, and
Measured Impact of Innovation and additional improvement in Experience with Innovation.
Simulating Different Strategies for Implementing Innovations in Schools
A number of simulations were done with the model to get an understanding of what is necessary
to successfully implement innovations. Initial conditions were established that made the model
simulate a “middle of the road” school system with Experience with Innovation that was
essentially neutral (a value of zero on a scale of -40 to +40). A reasonably benign history up to
that point had produced values of Teacher Motivation and Trust Between School and Community
of 0.5 on scales of zero to one. Because it is a fairly traditional school system, the Mode of
Student Evaluation and Structural Flexibility have values of zero and 0.1 respectively on scales
of zero to one.
The natural first step was to simply simulate the introduction of a large curriculum change by
itself. A large project was chosen that maximized all of the attributes (Breadth of Impact, Rote
Memorization vs Deeper Understanding, and Generality of Applicability) and was implemented
at the beginning of the simulation. Not surprisingly, it produced only a minimal Impact of
Innovation and almost no Measured Impact of Innovation at all. Experience with Innovation,
Teacher Motivation, and Trust Between School and Community all dropped precipitously.
Without attention to increasing Structural Flexibility, the Fraction A dopted was low and
Curriculum Innovations A dopted and Impact of Innovation were also low as a result. The small
Impact of Innovation that was produced also did not translate into a Measured Impact of
Innovation because nothing had been done to change the Mode of Student Evaluation. Clearly,
introducing a major curriculum innovation alone would not be successful.
The next step was to combine the major curriculum project with additional investments of time
in increasing Structural Flexibility and changing the Mode of Student Evaluation. (Both are
assumed to require significant numbers of teacher hours in addition to the curriculum
innovation.) All of these interventions were initiated together at the beginning of the simulation.
Rather than producing an improvement over the first simulation, the results were even worse if
that is possible. What happened? Trying to accomplish all of these changes at once left too little
time for doing any of them well or in a timely fashion. Curriculum Innovations A dopted
increased very slowly as a result. As the Level of Expected Impact increased, Curriculum
Innovations A dopted could not keep up. Experience with Innovation fell and caused Teacher
Motivation and Trust Between School and Community to fall as a result. With all of the changes
starting at once, the increases in Structural Flexibility and change in Mode of Student Evaluation
were too little and too late to really help. Once Trust Between School and Community began to
fall, it was also impossible to accomplish anything in these areas.
Delaying the curriculum project until after Structural Flexibility had increased and the Mode of
Student Evaluation had changed was a logical next step. The next simulation delayed the start of
the curriculum project by two years. The results were somewhat better, although not an
unqualified success. The Impact of Innovation and Measured Impact of Innovation exceeded the
Level of Expected Impact for a while, but leveled off while the Level of Expected Impact kept
rising. As a result, Experience with Innovation, shown in Figure 4, grew for a while and reached
a peak of 6 before declining. This modest, but temporary improvement helped to produce a
similarly temporary increase in Trust Between School and Community and contributed to a
modest increase in Teacher Motivation as well. Teacher Motivation also increased because the
change in Mode of Student Evaluation led to an increase in Perceived Need for Innovation.
Community Variables
6 pam
f anann
. ———
oa ‘N
yy \ NY
ol 1] Va N
0.4 — Th ‘\
0.4
0 15 30 45 60 75 90
Time (Month)
Experience with Innovation : S3 4 + + 4 + +
Trust Between School and Community: S3 2
Teacher Motivation : S3 3 3 3 3
Figure 4: Curriculum Innovation (Starting After Two Y ears) Together with Changes in Mode of
Student Evaluation and Structural Flexibility
How can the modest success achieved in this simulation be improved upon and result in a more
sustained impact? Will making more time available help? A simulation in which there was a
20% reduction in Tasks in Traditional Curriculum does, in fact produce better results in terms of
a graph of Experience with Innovation that reached a peak of 10 versus the peak of about 6 in the
previous simulation. (This simulation, as all of the remaining ones described in this paper,
includes changes in the Mode of Student Evaluation and increased Structural Flexibility as well
as the curriculum project starting two years after the beginning of the simulation.) However, as
in the previous simulation, Experience with Innovation declined toward the end as the Level of
Expected Impact again exceeded the Measured Impact of Innovation. Making more time
available by replacing the traditional curriculum rather than simply adding to it may contribute to
a successful strategy for implementing innovation, but does not by itself lead to a sustained
impact.
How about increasing the time teachers spend on professional development? Relationships in
Figures 1 and 2 suggest that a higher Level of Professional Development could have a favorable
effect by increasing Awareness of Innovation Elsewhere, a determinant of Teacher Motivation,
and increasing the A bility to Tailor Innovations to Students Learning Styles. Two simulations
were done in which the hours per month that teachers spend on professional development were
increased to 18 and 24 from the baseline value of 12. The results unfortunately were disastrous.
The benefits of spending more time on professional development were not realized since the
increased time on that activity took time away from curriculum development and kept the new
curriculum from being adopted in time to meet the Level of Expected Impact that developed.
Does this mean that professional development is a bad idea or a poor use of time? Of course not!
This result merely suggests that increased time on professional development must be combined
with other elements in order to contribute to a better strategy for implementing innovation.
What will work better? Increasing Flexibility Given Teachers Re: Innovation does appear to
make a big difference in producing a sustained effect. By increasing the Impact of Innovation
and Measured Impact of Innovation, this intervention sets several reinforcing loops in motion
that create a sustained effect rather than the temporary success seen in the earlier simulation as
displayed in Figure 4. Figure 5 displays the results of increasing Flexibility Given Teachers Re:
Innovation in addition to the interventions that were part of the earlier simulation. Rather than
reaching a peak and then declining, Experience with Innovation reaches a level of 15 and appears
to remain at that level. As a result, Trust Between School and Community achieves a level close
to one and stays at that level. Teacher Motivation achieves a higher level as well. (Note the
change in scale for Teacher Motivation and Experience with Innovation between Figures 4 and
5.)
Community Variables
20 2]
y
\
0 L377] Z
04 fal y
0.4 3 4
0 15 30 45 60 75 90
Time (Month)
Experience with Innovation : S8 + + + + + +
Trust Between School and Community : $8 2 2 2 2 2
Teacher Motivation : S8 3 3
Figure 5: Effects of Increasing Flexibility Given Teachers Re: Innovation
This result is encouraging, but can be improved upon. Two more simulations were done that
combined several of the interventions. One combined the increase in Flexibility Given Teachers
Re: Innovation with the 20% reduction in Tasks in Traditional Curriculum (in addition to the
other interventions contained in the earlier simulations). The second added to this already
comprehensive combination an increase in time spent on professional development (from the
baseline value of 12 hours per month to 18 hours per month).
Each of these enhancements produced somewhat better results. The first combination benefited
from the additional Hours Available for Curriculum Innovation made possible by the reduction in
Tasks in Traditional Curriculum. The Rate of Adoption was somewhat faster as a result and
Curriculum Innovations A dopted and Impact of Innovation grew more quickly. Experience with
Innovation reached and maintained a higher level. In the second combination, increased time on
professional development further improved the results instead of producing a disaster as it had
done in an earlier simulation. The reduction in Tasks in Traditional Curriculum had made the
necessary time available to do more professional development and still have enough time for
implementing the new curriculum. The higher Level of Professional Development that was
achieved helped to increase the Impact of Innovation by enhancing teachers’ Ability to Tailor
Innovations to Students Learning Styles
Another thing that was examined is the potential impact on innovation of other trends in
education such as “high stakes” testing being introduced in many states. A simulation that
assumed such testing might cause a 20% increase in Tasks in Traditional Curriculum as a result
of “teaching to the test” produced disastrous results. The additional time required for test
preparation left too little time available for curriculum innovation and the other changes required
to prepare the way for the new curriculum. The result suggests the need for a cautious approach
to planning innovation in the context of broader changes going on and for carefully replacing
traditional curriculum with new curriculum rather than simply trying to pile on more work.
Simulations were used to explore other issues as well. One focus of the work was the fit between
characteristics of schools and nature of innovations they can successfully implement. The same
set of interventions used in the simulation (S8) shown in Figure 5 were used with a school
system that was assumed to have lower initial levels of Teacher Motivation and Trust Between
School and Community (values of 0.25 rather than 0.5 on a zero to one scale). These lower
levels might be the result of some unfortunate events in a community’s past. As shown in Figure
6, the result of this simulation (INIT25) is significantly worse in terms of the Experience with
Innovation that is realized. The lower levels prevent the curriculum innovation and other
changes from getting off the ground and achieving the reinforcing effects produced in the
community with the better initial conditions. Rather than growing and achieving a stable level as
in the earlier simulation (S8), Experience with Innovation in this simulation (INIT25) reaches a
low peak and then drops off.
Does this mean that a community with such an unfortunate history cannot successfully undertake
curriculum innovation? Not necessarily. Such a community might instead have to approach
innovation in a different manner. Figure 6 also displays the result of another simulation
(INIT25TI) in which, instead of a major Systemwide Innovation initiated from the top down, the
emphasis is on Teacher Initiated Innovations. Experience with Innovation grows much more
slowly in this simulation, but achieves a moderately positive, apparently sustainable level rather
than dropping off as it did with the Systemwide Innovation.
10
Graph for Experience with Innovation
20
i a
9.6 aA
| 7] fa |
T27 >|
[47 “Ss
| IN
-0.8 La | aw
0 15 30 45 60 75 90
Time (Month)
Experience with Innovation : INIT25TI + + + + +
Experience with Innovation : INIT25
Experience with Innovation : S8
Figure 6: Simulations with Lower Teacher Motivation and Trust Between School and
Community
Conclusions, Next Steps
The work done so far, while tentative, suggests at least some rough “rules of thumb” for
implementing innovations in schools. The key insight is the interdependency of curriculum
innovations with other types of changes in areas such as student assessment and scheduling
systems. Curriculum innovation must also be planned in the context of changes going on in
response to external pressures such as “high stakes” testing. The simulation results also suggest
the importance of thinking carefully about replacing traditional curriculum with new material and
methods rather than simply adding on extra work. The last set of simulations reported on also
suggest the need for school systems to understand their strengths and weaknesses and implement
curriculum innovations accordingly. A final insight was the potential value of using System
Dynamics for examining the problems of innovation in schools and for other complex problems
schools are facing.
The next steps are to have the work described in this paper reviewed by a larger audience within
the System Dynamics community and among educators, especially those using System Dynamics
as part of their curriculum and/or to improve the management of their schools. Let us know what
you think. Does the model adequately represent the factors that affect the success of innovations
in schools? Is the response of various groups to innovations faithfully represented? How do
curriculum innovations interact with other things going on in schools? Does the model need to
be elaborated on in the area of student response to innovation? As indicated earlier, more
complete materials are available through the Creative Learning Exchange. Comments can be
communicated to the author via e-mail at GB Hirsch@ aol.com or at 7 Highgate Road Wayland,
Ma. 01778.
ve
