Could a collaborative database of OERs solve ULCs’ textbook problem?
University Language Centres (ULC), which provide language and academic skills support services to current and prospective university students, are major consumers of commercially published textbooks. Unlike other sections of the HE market, ULCs typically include a charge for textbooks in course fees then purchase the book on behalf of the students, thereby ensuring that all students have access to the same book. These textbooks are on the whole composed by the publishers with a great deal of forethought and skill; however, no coursebook is going to be able to cover all the needs of every student or provide the best tools required by every teacher to assist their students. These issues of cost and needs mean that there is a need for a more flexible method of selecting and distributing learning and teaching resources to students.
Open textbooks to the rescue?
Open textbooks (OT) are one method that has been used to overcome some of the shortcomings of traditional textbooks (Senack, 2015). These textbooks, which “[have] an open license that makes [them] free for anyone to use and change” (Center for Open Education, n.d.), are generally presented as being complete alternatives to standard commercial textbooks. That is they are fully-compiled sets of teaching and learning materials aimed at a particular type of student and designed to achieve a particular aim. Unlike commercial textbooks, the adaptation of OTs is not only allowed but encouraged. The Center for Open Education state they “believe the ability to make changes to an open textbook is integral to its definition as open” and even include a guide on how to modify an OT (Center for Open Education, 2016).
Does this make OTs the solution to the needs/cost problem? Alas, a commitment on the part of the producer to the right of users to adapt an OT does not mean the end users will find the process a straightforward or quick one. Indeed finding and then adapting an OT might prove more difficult and time-consuming than opting for a commercial textbook and supplementing it with some stand-alone handouts. Then there’s the fact that the cost of the resources necessary for ensuring that the OT found is suitable and that the adaptations are appropriate might make the adaption route little cheaper in the short term. Also, as Elsayed rightly notes, materials development can be seen as a ‘waste of time’ in institutions that do not give credit to staff that work on the development of OT materials.
A further problem for ULCs is that there are currently very few OTs available that attempt to address their particular needs. A search of the Open Textbook Library, for example, returned only a handful of books that might be considered even vaguely appropriate for ULCs.
These issues mean that OTs are not the obvious choice for ULCs and that they are more likely to either stick with commercially published textbooks or put something together from scratch.
A collaborative jigsaw
If not OTs, then what might solve the textbook problem? What I’m going to suggest is not so much a replacement for OTs but an extension of the idea. Whereas OTs are presented as off-the-shelf solutions, a more useful approach might be to offer customisable patterns with slots that can be filled from a repository of ‘pick and mix’ Open Educational Resources (OERs). That is, this approach would sit somewhere in between relatively fixed OTs and independent individual OERs with the end user being able to choice what to keep and replace but being guided in this process.
In practice this might mean that after searching for and finding a book that achieves the general aims that we have in mind, we would initially be presented with something very similar to a traditional OT. The difference would be that the book would be broken down into sections with clearly stated aims and intended learning outcomes (ILOs) and these sections would be further broken down into learning objects (LO), which also would have their aims and ILOs clearly stated. Each level of subdivision of the book would be linked to a set of alternative resources with matching aims and ILOs, as in Figure 1 below. The user would then more easily be able to take the parts that they liked and replace the parts that they didn’t.

Figure 1. Customisable OT pattern
The process would be rather like putting together a kind of jigsaw. We would choose the shape of the finished jigsaw we wanted and be presented with a finished picture. We would then, however, be able to take out the bits of the picture we didn’t like and fill in the gaps with appropriately ‘shaped’ pieces to produce something closer to what we’re looking for (à la Dave Gorman’s jigsaw at around 37’00”). Just as with doing a jigsaw, this is an endeavour that is almost certainly better done collaboratively.
Why might we collaborate?
ULCs typically spend £30 per student per term on textbooks for year-long pre-sessional courses. Whilst this is not a considerable outlay when compared with costs for teachers and rooms, it is money that would be better spent elsewhere both by departments and students. The savings made might just be enough to cover the costs necessary for completely new materials to be developed but are unlikely in the short term to cover the costs of developing the database and management system necessary to accomplish what is suggested. One reason for this is that the amount of resources necessary is likely to be great in terms of staff hours. But, even if sufficient time could be cleared in staffing timetable, it’s highly improbable that the average ULC is going to have the necessary IT skills to put together such a system considering that ULC staff generally have fairly low levels of IT skills.
How might we collaborate?
One solution, of course, might be to wait and hope that some group will identify the problem and set the system up for us. This has worked in the area of software where products such as Linux, Moodle and Firefox have been developed by Open Source Software (OSS) groups of volunteers that have developed high class solutions to widely-agreed problems. Some have suggested that such an approach would be a good idea for teaching and learning materials. However, as Mackie has pointed out such groupings have been far less successful in developing solutions to the more niche demands of institutions such as universities, and to overcome this problem, Community-Source Software (CSS) has been suggested as a solution.
In the CSS approach, a project is set up as a collaboration among a group of colleges that share a common need with each contributing resources to move the project forward. The participants form a virtual governing body to guide the project to match the aims identified at the start. However, the project remains open-source throughout allowing other parties to take what has been developed so far and fork off in another direction. This collaborative nature means that no one institution needs to commit all of the resources necessary, making it a more realistic endeavour, and means that each of the institutions can promote their individual needs, making the finished product more likely to satisfy the needs of a wide-range of users and thereby ensure its continued use.
Then what’s to stop it happening?
Despite the many possible benefits that such an endeavour might bring, there are a number of potential obstacles that need first to be considered and although none of them are trivial, I believe they are obstacles that can be overcome.
Two technical issues that need to be addressed before such a system could be considered user-friendly enough to be used by ULC staff have already been alluded to above: classifying the LOs so that they can be matched; and ensuring that the process of compiling the finished book is as painless as possible.
Ease of matching
Matching, by definition, involves making a decision on the similarity of objects in terms of one or more characteristics of the objects. This requires that decisions have been made on which characteristics are of importance and of how these characteristics are to be measured. In order to find a LO that was a match for the one we were interested in, we would first need to have data on the available LOs’ that described them in terms of their pedagogical potential. That is, a list of these data might include qualities that describe the object’s subject area, its intended audience or the level of interactivity it allows. But to find a suitable match we would also need to have access to data that described more temporal or physical attributes such as the expected time needed to complete the activity or the numbers of pages of A4 that the LO would occupy.
Working collaboratively across colleges could help to ease the process of deciding on which characteristics are of importance. The differing working conditions and demands of colleges might offer perspectives on the task. This is important since once LOs have been tagged with one set of characteristics any major changes in the nature of the set of characteristics can require a great deal of effort to reclassify and retag the objects.
Once the set of tags has been decided, the terminology used to describe each characteristic needs to be settled upon. So, for example, if one of the characteristics is ‘Difficulty’, agreement has to be reached as to how many levels of difficulty exist, what differentiates one level form another, and what these levels are to be called. Again bringing in the opinions of people from as wide a range of contexts as possible will hopefully ensure that any agreement results in a terminology that covers all eventualities.
Standards such as 1484.12.1–2002 — IEEE Standard for Learning Object Metadata have gone some way to defining a basic set of qualities that LOs might be expected to be tagged with. However, as the standard states it is merely a ‘base schema’ that can be used across a wide range of subject areas. Elements of such a standard could be used only if it was supplemented with more domain-specific tags. The IEEE’s standard, for example, contains a tag labelled ‘learning resource type’ but if the resource type was vocabulary development then a list of the specific lexical items practiced would seem to know when trying to match the LO, something which the standard doesn’t easily allow.
I’m not aware of any attempts to put together a basic ontology suitable for ULC LOs and my own efforts to do something similar so far have only strengthened my conviction that it would be more easily done as a collaborative project.
Ease of compiling
The jigsaw metaphor used above was used rather than the more usual Lego® metaphor because I agree with Wiley that seeing LOs as things that are ‘combinable with any other’ LO and that can ‘be assembled in any manner you choose’ obscures the complexities involved in putting LOs together to create ‘instructionally useful units’.
The proposal I have put forward here calls for OTs which are put together in units that are themselves made up of smaller subunits that might further be made up of subunits of increasing levels of granularity. This hierarchical structure means that particular LOs can only meaningfully be combined if they support the ILOs of the highest level. So if a unit is intended to develop a student’s knowledge of a particular lexical domain but one of the LOs used simply uses some of the lexical items in order to introduce a grammar point then it’s probably unlikely to be a good combination. Therefore, any collaboration would need to work carefully on a system of clearly stating the ILOs in the OT pattern initially presented to the user and of ensuring that the ontology used to classify the LOs is capable of representing ILOs.
A further issue that needs to be dealt with is one that anyone ever having put together a course pack will have experienced, and one alluded to by Elsayed, namely the confusion students have on coming across forward or backward references to material that exists in some other unit. To overcome this the collaboration needs to decide upon a set of rules for writing the LOs that, as much as possible, allows for such references to be either avoided or for them to be automatically replaced with an appropriate phrase or hyperlink to some other resource
Conclusion
Of course collaboration has its downsides: too many cooks may well spoil the broth if they do nothing but stand around debating it and never get around to cooking it. There will probably be an element of biting one’s lip at the early stages but the CSS route is based on the possibility of adaptation. Once a basic system has been set up, if one section of the ULC community believes that the system would be better developed in a particular way, they are free to fork off in that direction and see if the rest follows or not. The fact is that currently we don’t have any system for easily putting together customised OTs so the benefits could easily outweigh any potential obstacles if only we make that first step to setting up a system.
Anyone out there care to step this way?

