Archive for January, 2014

For some years now, universities and other educational institutions around the world have been using online learning platforms, also known as Learning Management Systems (LMSs) or Virtual Learning Environments (VLEs).Well-known versions of these include Blackboard  and Moodle. The latter is used by over 50% of higher education establishments in the UK (Dudeney & Hockly, How to Teach English with Technology Harlow, Essex: Pearson, 2007, p.53). These platforms allow course content – lectures, videos, activities, etc. – to be stored and delivered, and they allow institutions to modify courses to fit their needs. In addition, they usually have inbuilt mechanisms for assessment, tracking of learners, course administration and communication (email, chat, blogs, etc.). While these platforms can be used for courses that are delivered exclusively online, more commonly they are used to manage blended-learning courses (i.e. a mixture of online and face-to-face teaching). The platforms make the running of such courses relatively easy, as they bring together under one roof everything that the institution or teacher needs: ‘tools that have been designed to work together and have the same design ethos, both pedagogically and visually’ (Sharma & Barrett, Blended Learning Oxford: Macmillan, 2007, p.108).

The major ELT publishers all have their own LMSs, sometimes developed by themselves, sometimes developed in partnership with specialist companies. One of the most familiar, because it has been around for a long time, is the Macmillan English Campus. Campus offers both ready-made courses and a mix-and-match option drawing on the thousands of resources available (for grammar, vocabulary, pronunciation and language skills development). Other content can also be uploaded. The platform also offers automatic marking and mark recording, ready-made tests and messaging options.


In the last few years, the situation has changed rapidly. In May 2013, Knewton, the world’s leading adaptive learning technology provider, announced a partnership with Macmillan ‘to build next-generation English Language Learning and Teaching materials’. In September 2013, it was the turn of Cambridge University Press to sign their partnership with Knewton ‘to create personalized learning experiences in [their] industry-leading ELT digital products’. In both cases, Knewton’s adaptive learning technology will be integrated into the publisher’s learning platforms. Pearson, which is also in partnership with Knewton (but not for ELT products), has invested heavily in its MyLab products.

Exactly what will emerge from these new business partnerships and from the continuously evolving technology remains to be seen. The general picture is, however, clearer. We will see an increasing convergence of technologies (administrative systems, educational platforms, communication technologies, big data analytics and adaptive learning) into integrated systems. This will happen first in in-company training departments, universities and colleges of higher education. It is clear already that the ELT divisions of companies like Pearson and Macmillan are beginning to move away from their reliance on printed textbooks for adult learners. This was made graphically clear at the 2013 IATEFL conference in Liverpool when the Pearson exhibition stand had absolutely no books on it (although Pearson now acknowledge this was a ‘mistake). In my next post, I will make a number of more specific predictions about what is coming.

In order to understand more complex models of adaptive learning, it is necessary to take a temporary step sideways away from the world of language learning. Businesses have long used analytics – the analysis of data to find meaningful patterns – in insurance, banking and marketing. With the exponential growth in computer processing power and memory capacity, businesses now have access to volumes of data of almost unimaginable size. This is known as ‘big data’ and has been described as ‘a revolution that will transform how we live, work and think’ (Mayer-Schönberger & Cukier, ‘Big Data’, 2013). Frequently cited examples of the potential of big data are the success of Amazon to analyze and predict buying patterns and the use of big data analysis in Barack Obama’s 2012 presidential re-election. Business commentators are all singing the same song on the subject. This will be looked at again in later posts. For the time being, it is enough to be aware of the main message. ‘The high-performing organisation of the future will be one that places great value on data and analytical exploration’ (The Economist Intelligence Unit, ‘In Search of Insight and Foresight: Getting more out of big data’ 2013, p.15). ‘Almost no sphere of business activity will remain untouched by this movement,’ (McAfee & Brynjolfsson, ‘Big Data: The Management Revolution’, Harvard Business Review (October 2012), p. 65).

The Economist cover

With the growing bonds between business and education (another topic which will be explored later), it is unsurprising that language learning / teaching materials are rapidly going down the big data route. In comparison to what is now being developed for ELT, the data that is analyzed in the adaptive learning models I have described in an earlier post is very limited, and the algorithms used to shape the content are very simple.

The volume and variety of data and the speed of processing are now of an altogether different order. Jose Ferreira, CEO of Knewton, one of the biggest players in adaptive learning in ELT, spells out the kind of data that can be tapped[1]:

At Knewton, we divide educational data into five types: one pertaining to student identity and onboarding, and four student activity-based data sets that have the potential to improve learning outcomes. They’re listed below in order of how difficult they are to attain:

1) Identity Data: Who are you? Are you allowed to use this application? What admin rights do you have? What district are you in? How about demographic info?

2) User Interaction Data: User interaction data includes engagement metrics, click rate, page views, bounce rate, etc. These metrics have long been the cornerstone of internet optimization for consumer web companies, which use them to improve user experience and retention. This is the easiest to collect of the data sets that affect student outcomes. Everyone who creates an online app can and should get this for themselves.

3) Inferred Content Data: How well does a piece of content “perform” across a group, or for any one subgroup, of students? What measurable student proficiency gains result when a certain type of student interacts with a certain piece of content? How well does a question actually assess what it intends to? Efficacy data on instructional materials isn’t easy to generate — it requires algorithmically normed assessment items. However it’s possible now for even small companies to “norm” small quantities of items. (Years ago, before we developed more sophisticated methods of norming items at scale, Knewton did so using Amazon’s “Mechanical Turk” service.)

4) System-Wide Data: Rosters, grades, disciplinary records, and attendance information are all examples of system-wide data. Assuming you have permission (e.g. you’re a teacher or principal), this information is easy to acquire locally for a class or school. But it isn’t very helpful at small scale because there is so little of it on a per-student basis. At very large scale it becomes more useful, and inferences that may help inform system-wide recommendations can be teased out.

5) Inferred Student Data: Exactly what concepts does a student know, at exactly what percentile of proficiency? Was an incorrect answer due to a lack of proficiency, or forgetfulness, or distraction, or a poorly worded question, or something else altogether? What is the probability that a student will pass next week’s quiz, and what can she do right this moment to increase it?

Software of this kind keeps complex personal profiles, with millions of variables per student, on as many students as necessary. The more student profiles (and therefore students) that can be compared, the more useful the data is. Big players in this field, such as Knewton, are aiming for student numbers in the tens to hundreds of millions. Once data volume of this order is achieved, the ‘analytics’, or the algorithms that convert data into ‘actionable insights’ (J. Spring, ‘Education Networks’ (New York: Routledge, 2012), p.55) become much more reliable.

An integral part of adaptive learning programs, both the simple models already described and the much more complex systems that are currently under development, is an element of gamification. The term refers to the incorporation of points, levels (analogous to the levels in a typical computer game) and badges into the learning experience. In Duolingo, for example, users have a certain number of ‘lives’ that they can afford to lose without failing an exercise. In addition, they can compare their performance with that of other users, and they can win ‘lingots’, a kind of in-game currency which allows them to ‘buy’ lost ‘lives or to compensate for a day of inactivity.

duolingo lingots

Gamification and adaptive learning go together like hand in glove because of the data that is generated by the adaptive software (see the next post: Big data, analytics and adaptive learning). The whole thing is premised on comparing the performance of different students, so score cards and leader boards and so on are hardly surprising.

The idea behind this, in case it needs pointing out, is that it can make learning fun and, so, students will be more motivated to do the work, which seems more like play. It is a much hyped idea in education: eltjam referred to the ‘snowballing sexiness’ of the term. In an ELT context, most references to gamification are very positive. See, for example, eltjam’s blog post on the subject or Graham Stanley’s conference presentation on the subject. An excellent infographic summary of and advertisement for the benefits of gamification can be found at the Knewton website.

Not everyone, however, is so positive. Gamification has been described by some writers and researchers as the ‘pointsification’ of everything – the reductionist process of regarding all actions with points and increased personal scores (see, for example, Neil Selwyn, 2013, Distrusting Educational Technology, p.101). The motivation it may generate is clearly extrinsic, and this may not be a good long-term bet. Adults (myself included) get bored of gamification elements very quickly. For both adults and younger learners, once you’ve figured out how to play the system and get extra points (and there’s always a way of finding shortcuts to do this), interest can wane quickly. And once gamification becomes a standard feature of educational experiences (and not just English language learning), its novelty value will disappear.

‘Adaptive learning’ can mean slightly different things to different people. According to one provider of adaptive learning software (Smart Sparrow, it is ‘an online learning and teaching medium that uses an Intelligent Tutoring System to adapt online learning to the student’s level of knowledge. Adaptive eLearning provides students with customised educational content and the unique feedback that they need, when they need it.’ Essentially, it is software that analyzes the work that a student is doing online, and tailors further learning tasks to the individual learner’s needs (as analyzed by the software).

A relatively simple example of adaptive language learning is Duolingo, a free online service that currently offers seven languages, including English ( ), with over 10 million users in November 2013. Learners progress through a series of translation, dictation and multiple choice exercises that are organised into a ‘skill tree’ of vocabulary and grammar areas. Because translation plays such a central role, the program is only suitable for speakers of one of the languages on offer in combination with one of the other languages on offer. Duolingo’s own blog describes the approach in the following terms: ‘Every time you finish a Duolingo lesson, translation, test, or practice session, you provide valuable data about what you know and what you’re struggling with. Our system uses this info to plan future lessons and select translation tasks specifically for your skills and needs. Similar to how an online store uses your previous purchases to customize your shopping experience, Duolingo uses your learning history to customize your learning experience’ ( skilltree

Example of a ‘skill tree’ from

For anyone with a background in communicative language teaching, the experience can be slightly surreal. Examples of sentences that need to be translated include: The dog eats the bird, the boy has a cow, and the fly is eating bread. The system allows you to compete and communicate with other learners, and to win points and rewards (see ‘Gamification’ next post).

Duolingo describes its crowd-sourced, free, adaptive approach as ‘pretty unique’, but uniquely unique it is not. It is essentially a kind of memory trainer, and there are a number available on the market. One of the most well-known is Cerego’s cloud-based iKnow!, which describes itself as a ‘memory management platform’. Particularly strong in Japan, corporate and individual customers pay a monthly subscription to access its English, Chinese and Japanese language programs. A free trial of some of the products is available at  and I experimented with their ‘Erudite English’ program. This presented a series of words which included ‘defalcate’, ‘fleer’ and ‘kvetch’ through English-only definitions, followed by multiple choice and dictated gap-fill exercises. As with Duolingo, there seemed to be no obvious principle behind the choice of items, and example sentences included things like ‘Michael arrogates a slice of carrot cake, unbeknownst to his sister,’ or ‘She found a place in which to posit the flowerpot.’ Based on a user’s performance, Cerego’s algorithms decide which items will be presented, and select the frequency and timing of opportunities for review. The program can be accessed through ordinary computers, as well as iPhone and Android apps. The platform has been designed in such a way as to allow other content to be imported, and then presented and practised in a similar way.

In a similar vein, the Rosetta Stone software also uses spaced repetition to teach grammar and vocabulary. It describes its adaptive learning as ‘Adaptive Recall™’ According to their website, this provides review activities for each lesson ‘at intervals that are determined by your performance in that review. Exceed the program’s expectations for you and the review gets pushed out further. Fall short and you’ll see it sooner. The program gives you a likely date and automatically notifies you when it’s time to take the review again’. Rosetta Stone has won numerous awards and claims that over 20,000 educational institutions around the world have formed partnerships with them. These include the US military, the University of Barcelona and Harrogate Grammar school in the UK ( ).

Slightly more sophisticated than the memory-trainers described above is the GRE (the Graduate Record Examinations, a test for admission into many graduate schools in the US) online preparation program that is produced by Barron’s ( ). Although this is not an English language course, it provides a useful example of how simple adaptive learning programs can be taken a few steps further. At the time of writing, it is possible to do a free trial, and this gives a good taste of adaptive learning. Barron’s highlights the way that their software delivers individualized study programs: it is not, they say, a case of ‘one size fits all’. After entering the intended test date, the intended number of hours of study, and a simple self-evaluation of different reasoning skills, a diagnostic test completes the information required to set up a personalized ‘prep plan’. This determines the lessons you will be given. As you progress through the course, the ‘prep plan’ adapts to the work that you do, comparing your performance to other students who have taken the course. Measuring your progress and modifying your ‘skill profile’, the order of the lessons and the selection of the 1000+ practice questions can change.

There is a good chance that many readers will have only the haziest idea of what adaptive learning is. There is a much better chance that most English language teachers, especially those working in post-secondary education, will feel the impact of adaptive learning on their professional lives in the next few years. According to Time magazine, it is a ‘hot concept, embraced by education reformers‘, which is ‘poised to reshape education’[1]. According to the educational news website, Education Dive, there is ‘no hotter segment in ed tech right now’[2]. All the major ELT publishers are moving away from traditional printed coursebooks towards the digital delivery of courses that will contain adaptive learning elements. Their investments in the technology are colossal. Universities in many countries, especially the US, are moving in the same direction, again with huge investments. National and regional governments, intergovernmental organisations (such as UNESCO, the OECD, the EU and the World Bank), big business and hugely influential private foundations (such as the Bill and Melinda Gates Foundation) are all lined up in support of the moves towards the digital delivery of education, which (1) will inevitably involve elements of adaptive learning, and (2) will inevitably impact massively on the world of English language teaching.

The next 13 posts will, together, form a guide to adaptive learning in ELT.

1 Introduction

2 Simple models of adaptive learning

3 Gamification

4 Big data, analytics and adaptive learning

5 Platforms and more complex adaptive learning systems

6 The selling points of adaptive learning

7 Ten predictions for the future

8 Theory, research and practice
9 Neo liberalism and solutionism
10 Learn more