for Educational Research and Practice
Edmond Gaible, Michael Hannafin, David Merrill, Michael Spector,
Jan Visser, and David Wiley served as panelists for a discussion
on "Learning objects technology: Implications for educational
research and practice" that took place on the second day
of the Annual Meeting of the American
Educational Research Association (AERA), held from April
1 to 5, 2002, in New Orleans, Louisiana. The ideas behind the
panel discussion are explained in the attached approved conference
proposal authored by Yusra Laila Visser and Ray
J. Amirault, who coordinated and moderated the session.
Prior to the event, panelists had been asked to generate position
statements, summarizing their perspective on the advent of learning
objects and the opportunities and challenges that this new format
presents for learning specialists. In addition, they were asked
to generate questions that they would like other panelists to
respond to during the session. The following inputs were received
in response to this request (author names listed alphabetically):
The influence of educational approaches based on learning
objects is highly conditioned by context. Examinations of those
contexts and their histories may be instructive.
Military / corporate context
In military and corporate environments, courseware development
based on learning objects is tightly linked to the potential
cost savings and increased efficiencies of reuse. Development
of the Advanced Distributed Learning SCORM specification stems
directly from the 1997 Quadrennial Defense Review of U.S. armed
ADL the requirements of 2.5 billion armed forces personnel for
training-continuing throughout their military careers-and concludes
that given the high cost of multimedia development, of the more
than 30,000 delivered by the DoD without technological support
only a small percentage could cost-effectively be converted to
multimedia. A solution was seen in a reorientation toward reusability
(as well as persistence, discoverability, and modifiability)
as manifest in object orientation.
It should be noted, however, that the bulk of SCORM-related
development to date has focused on enabling mastery-level learning
and assessment for training purposes, with resources comprised
primarily of relatively static HTML- or XML-based courseware
and assessments. The key driver of ADL-based development is reusability,
not educational effectiveness.
In higher education, analysis of the current and potential
effects of learning objects is complicated by the activities
of parallel groups of developers. Substantial investment goes
to the creation of virtual universities and consortia (e.g.,
eCornell, Fathom), which provide expanded revenue opportunities
to institutions that are experiencing increasing costs that can
no longer be passed on to customers (i.e., students and their
families) or met by the expansion of on-campus enrollments. However,
university faculty, especially in schools that emphasize teaching
as opposed to research, are in some cases also encouraged to
develop Web-based resources independently, both to support their
on-campus courses and to support Web-based learning.
The MERLOT project (www.merlot.org/ ), hosted at the California
State University at Sonoma, links educators in over 20 state
universities through domain-specific communities. Faculty members
contribute learning objects that include web-based tutorials,
reference collections, articles, and dynamic media, as well as
interactive simulations. These objects are peer reviewed and
considered alongside publications in tenure decisions. Objects
are available for reuse by other faculty, however in this case
reuse is of secondary importance. The learning-object format,
in this environment, facilitates the emergence of faculty developers
by rewarding the contribution of discrete chunks of electronic
material, rather than comprehensive standalone resources, that
in many cases support textbooks, synchronous leader-based learning,
and other learning pathways. In cases such as this, learning-object-based
development is intended to enhance the quality, as opposed to
the profitability, of education.
In K12 education, impetus for learning objects is perhaps
the reverse of that of military/corporate training. Rather than
increasing the efficiency of vendor-based development, learning-object-based
development has the potential (only partially realized) to subvert
the dominant, private-sector resource-development paradigms.
Private-sector K12 development has split into three major modalities:
multimedia CD-ROMs, fee-based Internet portals, and advertising-based
portals. (In actuality, larger sites such as bigchalk.com combine
ads and fee-based access.) Structured for sales potential, rather
than utility-and in many cases designed for the consumer market
first and schools second-these modes foreclose participation
by broader populations.
As a format that is easily grasped and adopted by a wide range
of developers-including students and teachers-learning objects
can encourage broader participation in the creation of education
resources. Recent supporting developments, including software-authoring
tools (Agentsheets, Stagecast) and metadata models (Educational
Modeling Language) enhance the possibilities of object-formatted
development and for the user-developer in general.
Learning objects promote the exchange of resources and participation
in their development-activities that go against the private-sector
grain. Two simple example can be drawn from the Educational Object
Economy (www.eoe.org): First, the Turkish Ministry of Education
is funding localization of roughly 100 of the EOE's open-source
learning objects, in this case Java-based interactive simulations;
these localized objects will serve to "seed" development
of objects by Turkish educators working with private-sector developers.
Second, a privately developed web portal, Schoolarabia (www.schoolarabia.com),
is reviewing links to learning objects on the EOE site and integrating
these into its Arabic-language curriculum. In both cases, the
learning-object-based characteristics of EOE resources-discrete,
metadata-tagged, focused on specific curriculum items, and open
to a range of pedagogical approaches-makes them ideal for localization
into new and disparate educational contexts.
(It is reasonable, however, to suspect that object exchanges
such as the EOE have much greater relevance in countries that
as yet have limited experience with technology in education,
and that do not have mature education-software sectors.)
Within the K12 arena the most successful projects that expand
participation in learning-resource development-ThinkQuest and
I*EARN, for example-don't incorporate learning-object-based approaches
or metadata, in part because participation and the act of content-creation
are valued over the resources themselves. Efficiencies such as
those enabled by metadata, object exchanges such as the EOE,
component-based approaches such as ESCOT, and reusability as
an guiding precept may in fact run counter to the needs of learners
to experience the scaffolded building of knowledge.
And: Definitional questions, I believe, continue to dog discussion
of learning objects. Is a resource an object if it has learning-object
metadata (LOM) attached? Is it an object if it could have metadata
Imagine that on the edge of a schoolyard stands a cypress
tree identified by latitude and longitude, as well as LOM, with
teaching and learning supported by wireless, GPS, and handheld
technologies. At the tree, learners could have access to situated
research activities, simulations, dynamic media, diagrams, and
information, enabling learning to span botany, geology, geography
and even the natural history and traditional knowledge of the
region as it relates to the cypress. Is the tree a learning object?
Or are the learning objects only the resources connected to it?
I confess to the charge of thinking of learning objects, at least
primarily, in terms of the direct effect on development communities-vendors,
developers, teacher-developers-while seeing their influence of
learners and learning as resulting from these effects. I'm willing
to entertain all views in this matter, however. 
At the point, the Shareable Courseware Object Reference Model
(SCORM) aggregates the work of IEEE-LTSC, IMS, and AICC, integrates
XML, and adds specifications for an HTML-based run-time environment
to enable objects to be run in all SCORM-conforming Learning
Management Systems. The goal is for objects developed for use
by one branch of the services to be shared and re-purposed by
other branches, or by smaller units within a branch.
The Myths and Potentials
of Re-Usable Learning Objects
University of Georgia
The debate over the viability of "learning object economies,"
a concept advanced by Apple Computer roughly a decade ago, has
largely centered on three related issues: 1) Establishing and
implementing consistent meta-data standards; 2) Enabling access;
and 3) Establishing reasonable costing and cost recovery structures
to encourage and sustain participation in the learning object
economy. While progress has been made in each of these issues,
a more fundamental definition problem has brought many efforts
to a virtual standstill: What, in terms of size and function,
is a learning object?
Thus far, rather than becoming increasingly granular and usable
across many different contexts, learning objects have been progressively
enlarged into more familiar ideas of modules, units and courses.
While this may result in short-term efficiencies in the sharing
and re-using of intact objects, it is inherently constrained
by acceptance of the size and function of the object. In such
approaches, large numbers of otherwise discrete, addressable
objects, ranging from individual pictures to specific explanations
to segments and images within video, are collectively defined
as a single object. The potential for re-use of the multiple
components within a given learning object is rendered impossible
since their contents are neither tagged nor directly addressable.
So, to the extent potential learning objects are embedded into
larger units and strictly contextually dependent, we gain little
long-term efficiency in possible re-use in other, different contexts.
Long-term efficiencies from learning object technology, both
in the creation and implementation of learning environments,
depend on our ability to inventory and "warehouse"
libraries of re-usable, richly tagged objects. Rather than an
object's meaning being contextually fixed from their inception
and defined within this single context, the relevance and meaning
of any learning object is continually redefined by the changing
context of its use. A given object can be introduced across a
wide range of different contexts, by developers, teachers, and
students, and it's meaning reified situationally. Increasing
the granularity of complex instructional materials and media
resources can transform the ways learning systems are conceptualized,
the nature and role of media, and the tools used to both construct
and learning from learning environments.
In our conceptual-theoretical work (Hill & Hannafin, 2001),
we have developed a framework for resource-based learning based
upon granular notions of learning objects. Learning objects are
conceptualized as resources whose meaning is established differently
through different enabling contexts. Enabling contexts provide
situational boundaries in the form of problem statements, expectations,
issues and the like. In essence, they supply unique ways in which
the same given resource (object) assumes different meanings.
Resource-based approaches also scaffold learning in ways that
continue to reify an object's meaning according to the usage
context and tools that enable the user to act upon the problems,
issues and resources provided.
On the applied side, we teamed with industry researchers to
develop and test a resource-based EPSS known as TRIAD-Tactical
Readiness Instruction, Authoring & Delivery (Hannafin, Hill,
& McCarthy, 2002). Building on the resource-based learning
framework, and extending granularity concepts to account for
legacy data and multiple object use, we built a browser-based
system to aid naval personnel in creating and sharing tactics.
Through progressive queries, the system enabled naval officers
to generate and tag objects, construct documents dynamically
using defined objects, re-introduce the objects as organizers
as well as content to create and structure associated training,
and so forth. In effect, TRIAD is a support engine that uses
knowledge object technology to automate instructional design
and delivery as well as and document organization, creation and
Perhaps the most important lesson learned is that for learning
objects, size definitely matters, though its influence is often
negatively correlated with utility. As the unit size of objects
generated increased, our capacity to re-use the object in related
as well as different contexts declined. The problems associated
with creating, tagging and tracking learning objects across diverse
applications remains formidable, but the long-term payoffs warrant
the effort required. If we hope to find new best ways, we need
to resist the pressure to conform an otherwise transformative
technology to support largely traditional approaches. Inevitably,
the potential becomes marginalized and largely unexploited.
Hill, J., & Hannafin, M.J. (2001). Teaching and learning
in digital environments: The resurgence of resource-based learning.
Educational Technology Research and Development, 49(3),
Hannafin, M.J., Hill, J., & McCarthy, J. (2002). Designing
resource-based learning and performance support systems. In D.
Wiley (Ed.), The instructional use of learning objects
(pp. 99-129). Bloomington, IN: Association for Educational Communications
Brief Outline of my remarks as part of the Panel on Learning
Point one. As usually defined learning objects are
of little use to anyone. One definition indicated that a learning
object could be as small as a drop and as wide as the ocean.
If this is the case then everything is a learning object and
the concept has little or no value. My remarks will attempt to
define the relationship among media objects, knowledge objects,
instructional strategy objects, and learning objects. Media objects
are just that, small bits of text, graphic, video, audio. Knowledge
objects identify some instructional functions that these media
objects can play. When identified as a component of a knowledge
object the media object can now play an instructional role. However,
knowledge objects are still independent of instructional strategy.
When a knowledge object is combined with an instructional strategy
object then we have a learning object. We will also identify
knowledge bases as instantiated knowledge objects and knowledge
structures which identify the relationships among the components
of knowledge objects.
Our second point is that open-ended learning environments
so-called exploratory environments are often just indirect direct
instruction. That is, learners search for pieces of direct instruction
(learning objects?) that can teach them what they need to know
to solve the problem. This point will argue that for a student
to engage in self-instruction they need to analyze these pieces
of information to extract the knowledge components that they
need to solve their problem. They must then reassemble these
knowledge components in ways that provide the knowledge required
to solve their problem. Most students are ill equipped to provide
this analysis synthesis activity and are therefore left to search
for pieces of direct instruction. Knowledge objects, knowledge
bases, and knowledge structures provide a meta-mental model that
learners can use for this analysis and synthesis process. If
students are to learn the art of self-instruction then they must
be given the mental tools that will promote this self-instructional
Questions for other members of the panel.
1. Most meta data suggested to date does not related
directly to instruction. Is instructional meta data desirable?
What might it be? How can it be used in the formation of learning
2. Many seem to think that students already have the
necessary skills and that if we merely put them in these exploratory
environments they will learn. In my experience this is often
not the case. What then is required to promote self-instruction
in open-ended learning or exploratory environments?
3. Learning communities are all the rage. In my opinion
such learning communities are too often "pooled ignorance."
Or if they are not they consist of an inefficient search for
available direct instruction to provide the required knowledge
or skill. What might be required to provide resources that are
of real value to such communities? Is indexing this information
necessary in these situations? Why or why not?
1) A brief historical view of what has happened in
software engineering with regard to object technology - the gist
of this will be that much was promised, powerful methods and
tools were developed, but the goal of bringing computing to the
level of the problem solver unskilled in programming has not
occurred. Moreover, reuse of sophisticated and complex object
typically occurs only in restricted situations (e.g., within
an enterprise or a small community of skilled practitioners).
2) The user is generally left out of the reuse picture
- there are a number of user groups, including: (a) suppliers
of learning and training resources, (b) consumers of learning
systems; (c) system specialists; (d) instructional designers,
(e) teachers, and (f) learners. Various paradigms with regard
to learning objects leave one or more of these user groups out
of the picture. A simple-minded view intentionally omits the
in between groups based on the naïve belief that if objects
are properly tagged and made accessible by a knowledge vendor,
then learners will be able to find and use the appropriate resources,
especially if intelligent agents are operating in the background
to match learner characteristics and interests with various knowledge
resources. I hope to argue persuasively that this notion is too
simplistic since many learners require guidance in figuring out
what is worth learning and how best to go about such learning
- thus far beyond the capability of intelligent tutoring systems
and agent technology in all but well-defined and simple cases;
in short, there is still a need to involve teachers and instructional
designers, although what they do and need to know is likely to
change. In addition, since multiple consumer groups are involved,
there will continue to be a need to perform independent quality
reviews and evaluations of instructional resources - this will
also require instructional designers, teachers and content specialists.
3) The notion of widely accessible learning objects
is powerful and achievable. However, the ideals of those who
are making this technology possible may not be fully realized.
The irony here is similar to that cited by Naisbitt (1994) in
Global Paradox: the larger the world economy the more powerful
the smallest players become. In this case, the worry is that
the digital divide will actually widen although runs counter
to the views and preferences of those who have been working so
diligently on learning objects.
I will do my best to conclude with some positive ideas about
how best to realize the power and potential of learning objects.
Postscript (following the AERA session; added May 30,
In reflection upon the AERA discussion and the presentations
of the other panelists, my concerns have changed somewhat from
a concern about whether instructional designers might actually
begin integrating learning objects of one kind or another to
a concern about how this might be accomplished in terms of concrete
cases. As a result, I have developed a story-telling
case to promote some follow-up discussion to the session
in New Orleans. I would appreciate hearing reactions of any kind
(send email to firstname.lastname@example.org).
1) Wiley (2000) argues that the definition of learning
objects by the Learning Technology Standards Committee (LTSC)
is too broad and proposes to limit the use of the term to digital
objects. This makes sense from a practical point of view. However,
it remains of value to consider continuity of practice among
instructors, and designers of instruction, to have recourse to
reusable building blocks - whether digital or not. Some do so
in mindful ways; others don't.
Recall Paul Simon's lyrics:
"The problem is all inside your head," she said
The answer is easy if you take it logically
I'd like to help you in your struggle to be free
There must be fifty ways to leave your lover....
If there are fifty ways to leave your lover, there must be
at least a thousand ways to use a pair of pliers, and an almost
infinite number of ways to use a learning object, considering
the complexity of learning processes. The point is not, in my
view, that we should look at instructional building blocks as
atoms or molecules that have different bonding characteristics
(Wiley's metaphor) as opposed to pieces out of a LEGO box. Just
as I can use a LEGO block in many different ways (for instance
by cutting it into pieces, melting it, hammering a nail into
it, gluing a magnet onto it, etc.), so can I use a learning object
in many different ways. The problem is with the mindset of the
user ("it's all inside your head"), or more so even
with the absence or presence of mind in the user. If the user
is an unimaginative executer of procedure, stacking one block
on top of the other, one gets something totally different than
when the user uses his/her mind and visualizes, at every moment
a new block is to be added, a whole new range of possibilities
regarding how to continue. This difference is well expressed
in the comparison of the two stone walls depicted below: procedural
behavior on the left and mindful behavior on the right.
The application of mind rather than procedure will eventually
lead to constructions - such as shown in the picture below -
that satisfy our sense of beauty and thus invite us to interact
with them. Well crafted instruction is like such a well constructed
building. It is a pleasure to look at and a joy to enter into.
2) The trouble with learning objects resides not so
much in what they are - mere objects - but rather with the implicit
ideas behind their supposed usage, namely as ready?made off-the
shelf implants into the set of conditions that prompt and guide
human learning. Particularly troublesome is the narrow perspective
on learning (acquisition of "knowledge as a thing"
from some source) that frequently underlies these ideas as well
as the implicit assumption that learning objects are neutral
with regard to their context.
3) Wiley (2000) comments: "The problem was that
no instructional design information was included in the metadata
specified by the current version of the Learning Objects Metadata
Working Group standard" (p. 11). This is certainly a problem.
However, even if such information had been included, it would
still depend on the particular perspective on instructional design
(either viewed as dealing with a closed system or as applied
to an open system that interacts with its environment), whether
one would end up with an attitude towards using learning objects
that would do justice to the complexity and non-linearity of
human learning. From the perspective of instruction as a closed
system, the instructional design metadata that accompany a learning
object will be perceived as fixed and rigid, calling for the
"correct" procedure to be applied, inserting the object
in the one or few appropriate ways prescribed by the metadata,
leaving no place for the imagination of the designer/instructor.
4) During a recent presentation by a major European
player in the area of e-learning development, in which reference
was made to the use of learning objects in the development of
"e-learning solutions," senior officials present embarrassed
themselves by not being able to respond adequately to the following
question: "Given a number of objects all conforming to the
SCORM standards, when should they be considered learning objects
and when not?"
5) I conjecture that the most dangerous learning objects
are the ones of medium size. They are big enough to be seen as
capable of provoking the execution of a particular learning task.
Given the tendency to "make life easy," designers/instructors
will be tempted to just plug them in without further thinking
and, worse even, without having recourse to their most human
resource: imagination. If learning objects are small enough,
they can only be used as nuts and bolts, useful things to build
larger structures, using imagination and insight while doing
so. If learning objects are large enough, they will represent
a level of complexity concomitant with the complexity of the
learning process, requiring imagination on the part of the different
actors in the learning environment to interact with them adequately.
Examples of the nuts-and-bolts type of learning objects are,
in my view, the java applets available in the collaborative environment
of the Educational Object Economy (http://www.eoe.org). Examples
at the other end of the spectrum are such Web sites as http://www.reusability.org,
http://tip.psychology.org, and http://www.learndev.org.
6) The report by the Delors commission to UNESCO (Delors
et al., 1996) calls for education to move beyond the level of
acquisition of skills and knowledge. It defines "learning
to know," "learning to do," "learning to
live together" and "learning to be" as the four
essential pillars of education for the twenty-first century.
Clearly, such a thesis moves education, and thus human learning,
out of the realm of useful commodities and into the sphere of
essential social phenomena that will determine our chances to
co-habitate the planet while aspiring to "attain the ideals
of peace, freedom and social justice" (p. 13). Such education,
which is lifelong, is by nature of the social aspirations it
undergirds, served by keeping an eye on the whole while exploring
the detail. Much of traditional educational practice, however,
is based on a strong preoccupation with detail (disciplines,
specialized knowledge) as opposed to the whole, the latter concern
requiring a transdisciplinary vision (Nicolescu, 2002). Morin
(1999) thus refers to the "second scientific revolution"
(p. 28), which started in the nineteen-sixties (the first one
having upset the physics of the very small during the beginning
of the twenty-first century), the focus of which is - like in
the case of ecology, the Earth sciences, and cosmology - on complex
systems that by virtue of their complex adaptivity can serve
as organizing wholes. While such a vision is not at all contrary
to the possibility of using small information packages in processes
that prompt and facilitate human learning, it definitely calls
for a toning down of the discourse that touts learning objects
as the powerful means through which learning will make its next
Question for the panel:
What do we think are the implications for the mindset of instructional
designers of emphasizing the prescriptive nature of design theory?
The reasons why I'm asking the question are the following
The philosophy that underscores the use of learning objects
is firmly embedded in the perspective of prescription.
I have always found the idea of "prescriptive theory"
difficult to swallow and doubt whether there is any other field
of scientific endeavor where people have found the need (or convenience)
to introduce the term. Certainly, as a theoretical physicist
I'm not aware of any theory concerning physical phenomena having
been declared "prescriptive." Even in the medical profession,
where a lot of prescription is going on, I don't think anyone
speaks of a presriptive theory.
Theories are broad conceptual frameworks inductively derived
from a limited set of facts. They are the broad pictures of what
those individual facts that we know seem to fit into. Once developed,
theories can be used deductively to predict certain things. Such
predictions can then be checked to either confirm or falsify
the theory. The idea of a theory being "prescriptive"
- as opposed to "descriptive" - seems to me somewhat
odd in the above perspective. The term "prescriptive theory"
suggests we are sure of what we know (even if we call it probabilistic).
This takes much of the fun out of the meaning of the very word
The emphasis on presription leads designers and instructors
to view their work in a mechanical fashion, not having recourse
to their imagination. Skinner, whom we have since condemned,
addressed in a seminal article in 1954 the question of "the
science of learning and the art of teaching" (my emphasis).
We have much progressed since 1954 in our understanding of learning.
Unfortunately, it hasn't apparently inspired the instructional
design community to start looking at its vocation as an art,
rather than a procedurally executed technology.
Delors, J., Al Mufti, I., Amagi, I., Carneiro, R., Chung,
F., Geremek, B., Gorham, W., Kornhauser, A., Manley, M., Padrón
Quero, M., Savané, M-A., Singh, K., Stavenhagen, R., Suhr
M.W. & Zhou N. (1996). Learning: The treasure within. Report
to UNESCO of the International Commission on Education for the
Twenty-first Century. Paris, France: UNESCO.
Morin, E. (1999). La tête bien faite: Repenser la réforme,
réformer la pensée [The well-made head: Rethinking
reform, reforming thinking]. Paris, France: Seuil.
Nicolescu, B. (2002). Manifesto of Transdisciplinarity. New
York: State University of New York (SUNY) Press.
Skinner, B.F. (1954). The science of learning and the art
of teaching. Harvard Educational Review, 24(2), 86-97.
Wiley, D. A. (2000). Connecting learning objects to instructional
design theory: A definition, a metaphor, and a taxonomy. In D.
A. Wiley (Ed.), The Instructional Use of Learning Objects: Online
Version [Online]. Available: http://reusability.org/read/chapters/wiley.doc
[March 10, 2002].
Photo credits: Jan Visser
Pictures were taken in Eyragues, a small village in the Bouches-du-Rhône
in southern France.
1) Many learning technology standards (SCORM, IMS,
LTSC) seem to assume that next-gen technology-enhanced instruction
will be comprised of completely reusable components. As I have
discussed elsewhere (e.g. http://wiley.ed.usu.edu/docs/cisco_rlo.html)
this assumption promotes the simple memorization of trivial,
disconnected facts and prevents instruction and assessment of
integrated performances which are valuable in the real world.
This has to be addressed.
2) There has been a clamor for an archive of peer-reviews
of learning objects and other instructional materials, so that
a busy teacher could drop in and find effective resources quickly.
Unfortunately, most of what we currently know about evaluation
tells us that the best evaluations are those most closely tied
to the context of use of the evaluation data. This is problematic
because evaluations of a learning object (supposedly useful in
a variety of instructional contexts) must focus on a single instructional
context (to be optimally useful), thus telling the teacher nothing
about how the object will function in her own instructional context
(unless she happens to be planning to deploy the object in exactly
the same context as the reviewer foresaw). This has to be addressed.
3) Many of the most innovative and instructionally
effective uses of reusable instructional content happen outside
of LMSs, without the support of metadata, sequencing, or other
specifications, and without any other tools that instructional
technology would generally claim as its own. The new scalable
learning architecture is the massive self-organizing online community,
capable of responding to queries without databases, providing
infinitely scalable human-to-human learning support, and already
fabulously popular online. Chances are you're already part of
one, and just not appreciating the significant learning going
on in that informal environment. (Yes, Merrill and I will go
round and round about this :) It's time we began appreciating
informal learning environments as valuable instructional settings,
and letting functioning, successful practice drive our theory.