This all seems well and good but how can this be achieved in an era of large student/staff ratios and reports in the literature that students do not read the feedback but only pay attention to the mark.  Feedback, as a term, is also debated and researchers are leaning towards the notion of FeedForward, which will prompt students to reflect on the tutor’s appraisal of their work and take the ideas forward into their next assignment.  However the evidence to date is that tutors are working very hard at promoting reflection but is the advice being taken seriously and used by students in subsequent course assignments?

One of the issues researchers seem to agree upon is the timeliness of the feedback. However receiving constructive comments at the end of an assignment is perhaps not as useful as we think and it is during the construction, or co-construction, of an assignment that the student needs help. There has been a long tradition at the UK’s Open University of students being able to ring their tutor when they have difficulty and speak to them, usually in the evening, when they themselves are not at work and the tutor too is not teaching.  This system has worked very well but ideally all students, whether they are studying a distance course or attending a face to face university, would wish to have a tutor respond promptly to their queries when they are working late at night and so the feedback is timely and more meaningful to them.

There is a commercial company based in the United States called “Smarthinking”  which is trying to address ‘feedback on demand for students’. It provides just what it states on the tin which is on-demand, online tutoring.  It has a bank of tutors who support students studying conventional courses. It is often a source of extra help paid for by the University or College which has been outsourced to the Smarthinking company.  The service is available 24 hours a day, 7 days a week and is now being used by a number of distance universities in Australia.  The grades for Smarthinking students were significantly higher according to a study undertaken by Jane Calfee at the Kapi’olani Community College but what is important here is that tutors are commenting on drafts of assignments. The university or college tutors can see the data produced by the company on their student’s progress.  They can read the tutorial sessions and can then spend their tutorial sessions on dealing with student problems.  Hence a change in pedagogy has occurred.  Is this the future of digital dialogues?

As at any conference there is always a hot topic which was assessment for learning and how to provide useful feedback was on a lot of peoples’ minds. The term feedback itself was in dispute with some researchers favouring the term feed forward (Dai Hounsell)  but perhaps what was more interesting was not the terminology but the appropriateness and timeliness of the responses given by tutors to the students. I was delighted to see that researchers such as Margaret Price, Karen Handley and Berry O’Donovan in their paper on feedback, together with Sue Bloxham and Liz Campbell from the University of Cumbria, were emphasising that feedback should be seen as embedded within an ongoing dialogue.  Their presentations are available in the book of abstracts. This notion of promoting a dialogue between student and tutor is one which Stuart Watt  and myself have been progressing with our development of OpenMentor   and Open Comment.  Both these systems provide automated feedback to the user and lets them know what they didn’t know! 

The three Plenary Lectures provided three different vignettes into current thinking about assessment. Eckhard Klieme  discussed a number of pertinent issues that impinged upon educational measurement which is of particular interest to all those involved in PISA testing. His presentation was entitled ‘Assessment, grading and instruction: Understanding the context of educational measurement’.  Of particular interest was Ruth Leitch’s topic which informed us about children’s rights, as enshrined in European law, with respect to the assessment process.  It was Dylan Wiliam’s Plenary which closed the conference and was aptly named ‘When is assessment learning-oriented?’  He interestingly challenged the notion of formative assessment per se and has devised a new definition with Paul Black which they hope will be published in a paper in 2009.  They state:

 “An assessment functions formatively when evidence about student achievement elicited by the assessment is interpreted and used to make decisions about the next steps in instruction that are likely to be better, or better founded, than the decisions that would have been made in the absence of that evidence.”

 The emphasis here of course is upon instruction but shouldn’t a definition include a subsequent action from the student?

This conference did not solely address forms of electronic feedback and assessment but participants‘ papers which focused in this area were invited to submit their contributions to a special issue of the British Journal of Educational Technology  which I am editing.  It should be available in March 2009, Volume 40, No. 2. So watch this space!

Trevor Barker’s paper ‘Computer Adaptive Testing in Higher Education: The Validity and Reliability of the Approach’ was based on a six year study into the design and evaluation of a  computer-adaptive test (CAT) which had been used with Computer Science undergraduates at the University of Hertfordshire.  The statistical findings reveal that the CAT was able to match the level of difficulty of the test to the ability of the students.  More importantly the validity and reliability of the CAT approach stood up well against other forms of computer assisted assignments. This is an important finding for developers especially those concerned with PISA   and TIMSS  testing, where there are real concerns about the validity and reliability of all test items but more especially questions administered via an electronic medium.  

Another group of researchers led by Pete Thomas, have been investigating the ‘Automatic Assessment of Sequence Diagrams’  They have found that the computer marks the students’ entity-relationship diagrams more reliably than the tutors!  The significance of this group’s system is that it not only detects the students’ syntax errors, but also provides them with constructive feedback which is also in the form of a diagram.  Therefore output to the student matches input. I believe this to be a really important point with respect to feedback which indeed must be meaningful to the student and should need little decoding if the advice is to be acted upon.

I must confess to having a vested interest in the third paper as it was written by myself and Stuart Watt and tells the story of ‘Open Comment’, an automatic formative assessment guidance tool for History students.   This work addresses one of the issues raised  by myself and Andrew Brasher when we devised the JISC Roadmap for e-assessment and that was to address the challenge of providing students with interactive tasks that supported more free text entry and provided the students with immediate feedback.  Although Open Comment was designed to be used within the Moodle environment, it has an open and flexible framework and there should be no significant difficulties in adapting or embedding it into other formative assessment systems.

The most prestigious paper that I want to draw your attention to is the one authored by Alison Fowler from the Computing Department of the University of Kent.  It was entitled ‘Providing Effective Feedback on Whole-Phrase Input in Computer-Assisted Language Learning’ . Aliy has built the LISC system which is language independent and which does not rely on parsing errors in the input text to provide the student with effective feedback. Instead the novel approach of a sequence comparison method was used, which has not been applied to language learning in such a systematic and rigorous way before, although this method has proved successful in the fields of biology and chemistry with respect to gas chromatography.

It was my privilege and pleasure to present Aliy with a bottle of champagne for the ‘Best Paper’ Award at the CAA Conference.  She has a free delegate place for next year’s Conference and I look forward to hearing about further developments in the LISC system. Well done Aliy!

 Aliy Fowler receiving from Denise Whitelock the ‘Best Paper’ Award at the 12th CAA Conference, Loughborough, 2008

   Here’s a tool  Firefox3 ,that was launched on Tuesday June 17, should help you check it out  more easily. Why’s that you ask. Well, its a speedy  animal  with one-click bookmarking with  a great zoom in and out facility  which is what you want from your internet browser but that’s not all. What else would you put on your wish list?  I certainly would like help to revisit sites  when I have forgotten the address.

The ‘Awesome Bar’ helps you do just that by matching the words in your search entry with words from the websites that you have visited before. I guess the usefulness of this facility will improve over time as will its accuracy when a better picture is built of where you go to check it out!

Firefox in space?

Creating opportunities for dialogue is an essential part of the learning process and is widely accepted as a suitable activity in the constructivist theory world.  However formal assessment has consistently failed to follow through with this more recent thinking because it squares the desire for improved constructivist learning against the demand for institutional and external reliability and accountability. So the official “checking it out” procedures have not kept apace or have they?

I was very interested to discover this week the notion of the “peer exam” which was developed by Vera John-Steiner . With the “peer exam”, (which is more reminiscent of androgogy than pedagogy), the students choose a topic they wish to explore; which is meaningful to themselves and their colleagues. Two to four students can take part in this process. They work out the questions they will use and prepare for the exam by reading around the topic and talking to one another. The students tape the examination session and write up the process as well as the results of their deliberations. Although this type of examination feels uncomfortable to begin with, it has had a profound effect on the students, some of whom have adapted this process to their own teaching. John-Steiner based this “peer exam” on Vygotsky’s theory of learning which emphasises the social nature of knowledge acquisition.  So what can we take from this?  Perhaps it’s best to check it out together?

Checking it out together: preparing for a “peer exam”

Presenters were asked to think about how we could exploit technology to sustain lifelong learning and how we can support learners changing needs in the future.  Quite a task, don’t you think?  So I got out my crystal ball and started to consider about the sorts of issues that I am dealing with in e-assessment and then thought it was better to move onto the social networking tools and elearning communication systems today that we can learn from to meet the challenges of tomorrow? 

 I would suggest that one of the common activities is that of ‘checking out stuff’.  So what do I mean by that?

Let’s take the example of Facebook. Users can check out the following:

·         Checking out whether their friends are online or what they are doing.  Twitter too assists with this functionality

·         Arranging meetings/dates

·         Seeking advice about places to visit, restaurants etc.  There are applications such as Local Picks Tripadvisor to assist

·         Appraising reading lists and looking for a good read  using Visual Bookshelf

If folk continue to ‘check out’ they do so because there is a response and new tools keep appearing in places like Facebook for friends and colleagues to record their opinions about a number of phenomena which are open to others. 

This checking out also takes place in learning and a number of models have been developed which explore conversational relationships in education e.g. Laurillard (1993) and Pask (1976)  (Unfortunately I cannot find a free copy of Pask’s 1976 paper online.  Can anyone help with this?)

Technology can play an essential role here in providing the feedback which facilitates the exploration of new roles in this new ‘check it out’ arena. In fact electronic feedback systems that support student interactivity have proved to be successful in providing audio feedback that explains the reasoning behind correct and incorrect choices for answers to a set of ‘check it out’ questions.  I have found this strategy to be very useful in the Open University’s Science Foundation Course (S103) where all this information would not be provided in a face to face tutorials.  This is because tutors would only have time to explain the correct solution to a problem and go through the workings of that solution but would not necessarily describe why all the invalid incorrect answers were wrong.

Therefore I think we should exploit the technology not only to support just in time check out for learners but use the feedback in these systems to help shape the users  as independent thinkers, making their own judgements and decisions about their learning process. This information will in turn assist them to make choices about their own futures. One way to achieve this outcome is to build new feedback systems which will facilitate novel roles for learners which fit within a constructivist perspective and meet the demands of The Leitch Report (2006) which proposes  a 45% expansion in Higher Education by 2020.  Dearing, in his speech to the Society of Research into Higher Education’s Governing Council, February 2008 believed this figure to be unrealistically high but stressed the need for such an expansion in order to achieve the same goals he stressed in his 1997 report which were to:

·         ‘serve the needs of an adaptable, knowledge based economy at local, national and regional levels

·         play a major role in shaping a democratic, civilised, inclusive society’  (Dearing, 1997)

New feedback systems are therefore required to assist learners to make choices about their own futures and how these sit within a challenging and changing economy.

Moving Forward

The race has been on in the United States since the mid-1960s  to decode free text entry of students’ work for automatic essay grading.  E-rater (ETS)  and the Intelligent Essay Assessor (IEA) currently in use. The latter uses Latent Semantic Analysis (LSA).  The strength of both these systems is that they can be used to give constructive feedback. The strength of IEA is that it can be used to give constructive feedback. IEA has been described as a psychologically sound

system as it is held by some of its proponents to be a theory of language use in people, and indeed, it does seem to match several significant psycholinguistic effects (e.g. Landauer et al., 1997) 

The problems associated with the evaluation of free text entry responses have also been concentrating the minds of myself and a group of researchers at The Robert Gordon University. We are currently working on ”Open Comment”  which is an open source text recognition question type for Moodle and was built to facilitate free text entry for students answering History and Philosophy questions for a series of formative assessments.  Open Comment implements a constructivist and supportive approach to providing feedback, based on the following key principles:

• There is no ‘right’ answer
• Any response by a learner will generate a mixture of supportive comments and suggestions for improvement
• The suggestions for improvement are selected on the basis of identifying markers extracted from the text
• The mechanisms for identifying markers are flexible, and do not need to depend on keywords in the text itself

The current implementation of Open Comment includes two classifiers, one based on regular expressions (text patterns) and the other using support vector machines, a classifier based on machine learning technology. This is an approach which is proving successful and offers one path of investigation for a ‘lifelong learning check it out’ system.

Any developments in this area will need to adopt an interdisciplinary approach to the research and this strategy will be required to tackle the problems associated with producing new feedback tools for learners. Learners are often checking things out due to an all pervasive assessment regime, and changes in feedback  for learning systems require changes at levels that range from the minutiae of diagnosing student progress on a specific task to influencing political decision making. We will certainly need a number of partners with expertise in areas such as:

·         The design of information /feedback tools

·         Latent semantic analysis and associated areas of expertise

·         Innovative e-assessment

·         Educational evaluation and theory

·         Mobile learning

·         Curriculum design and skills based courses in Higher Education

·         User modelling

·         Open source development of tools for enhancing learning

But what do you think?  Where can we go with this? Perhaps the ideal would be when the student uses the system has confidence in it and believes the feedback will help them progress and do better!