Teaching and Learning Forum 97 [ Contents ]
Peer assessment as a tool for learning
Edith Cowan University
This paper describes a set of strategies for peer assessment introduced into a teacher education course, in a attempt to lessen lecturer workloads associated with assignment marking while improving students learning. The particular aim of the strategies used was the development of students' ability to analyse academic expectations, to give and receive critical feedback and to learn through reflection. During the first year in which peer assessment was introduced lecturer workloads were not reduced but potential for this outcome is predicted, once the work of establishing new procedures and approaches is complete and students develop confidence in the new course aims and methodology. Anecdotal evidence supports the idea that learning outcomes for students can be improved by their active involvement in peer assessment, but not all students enjoy working in this way and some find the approach does not match their beliefs about learning well.
As a lecturer in a large Faculty of Education I faced many challenges in trying to facilitate the development of higher order learning in students: two assessment issues confronted me with particular force:
My twin goals, therefore, became to find ways of giving feedback which would be;
- Given the large number of students for whom I was responsible and the extensive volume of work submitted for assessment, I found the workload associated with marking to be overwhelming. In order to evaluate work accurately and fairly and to give students what I felt to be adequate feedback, I found myself working routinely beyond 70 hours per week. Since I believed that feedback was critical to the development of student learning, it became clear that I had to find more efficient ways of providing students with relevant feedback.
- Although I gave detailed (and I thought clear ) written feedback to students about their individual assignments, I found that subsequent work submitted showed the same pattern of errors. I investigated institutional records which contained formal tutor feedback on student work and found that it was not uncommon to find criticisms made of assignments done by a student in their first year of study still being made about work submitted in year three. I reasoned that either the students were not reading and understanding the feedback or they did not wish to improve their performance. As my goal was to constantly improve the learning and achievements of all the students, it became apparent that I needed to increase the active use of feedback.
- efficient (in terms of lecturer input)
- effective (in terms of improving students learning)
It seemed appropriate to try and obtain some insights from students on the matter. I began by sharing my observations with them in their seminar groups and then collected their thoughts, explanations and interpretations of the matter simply by engaging them in individual and group discussion. As with much action research, my focus was on solving an immediate professional problem and although valuable anecdotal evidence was available, I was too inexperienced at the time to realise the potential of the data and it was not formally recorded in any way. From the discussions I did, however, note a range of interesting points. Several factors seemed to limit students ability to use feedback effectively:
- Time: there was not enough time to reflect upon work both during its preparation and after its return
- Quality and quantity of feedback: Staff did not provide feedback which was consistently detailed, relevant to improvement and understandable
- Attitudes to self: Some students held the belief that innate ability was a dominant factor and that learning, or working hard would not influence the outcome
- Rewards or penalties for responding (or not ) to feedback were not apparent to students
- Knowledge about effective study was underdeveloped in some students
The teaching team met to review the problem. As a result of their debate the course programme for the following year was adapted to allow for a range of changes which included the introduction of peer assessment. This initially featured:
- whole class, group and individual assessment of student performance and contribution during seminars / tutorials
- peer appraisal of lecture notes, required reflections on lectures and directed reading tasks
- peer assessment of written assignments.
Description of peer assessment strategies and their implementation
1) Peer assessment of student performance and
Expectations of student participation and performance in specific learning situations were discussed regularly and guidelines negotiated and agreed between the lecturer and the student group. An aspect of behaviour would then be selected as a focus during a teaching/learning activity, and criteria for and methods of assessing that behaviour would be agreed. One or two students would then be elected to observe their peers throughout the teaching/learning session, tally relevant information and give the group feedback at the end of the session.
contribution to seminars and tutorials
This strategy was simple to implement. It was easy to explain, could be applied with great flexibility and did not require any significant resources or time. Some students were initially uncomfortable with the idea of being observed and were concerned that they might be shown up in an unfavourable light. The problem was resolved by the negotiation of procedural rules which focussed feedback on information rather than judgements and by an agreed commitment towards group rather than individual responsibility for achieving required changes, along with an acceptance that different perspectives, needs and values exist and may need to be accommodated. A variety of activities were also introduced to help the student group establish more trusting relationships. The group were quick to adopt the cycle of continuous improvement through goal setting and feedback as a standard procedure to accompany learning activities and were regularly observed establishing such frameworks even when the lecturer had not explicitly required them.
One example of a successful outcome of this strategy was a positive change in the gender balance of contributions to discussions. A group agreed that as a general principle males and females should contribute equally to discussions. A simple tick sheet of contribution by gender was kept over a 3 week period, with the statistics fed back to the group at the end of each session. By the third week the group had altered their behaviour from male dominated discussion to more evenly distributed discussion.
The discussions about expectations were particularly valuable. It became clear to both lecturers and students that expectations were not always shared and that the meanings attached to the terms used to describe expectations could vary greatly. This highlighted the need for careful analysis of the teaching /learning situation in order to identify relevant criteria for evaluation. Academic staff became more aware of the need to communicate about their expectations both within teaching/learning situations and in the context of formal assessments. Although criteria for assessment had previously been published to students as a matter of routine, the care and attention given to setting and describing criteria increased noticeably, as did the willingness to include opportunities for negotiated criteria matched to individual student needs and values. Likewise students became empowered to identify situations with a potential for mismatched assumptions and to seek clarification .
2) Peer assessment of lecture notes, required reflections
Lecturers had an implicit expectation that students would take notes during a lecture or workshop, reflect on the material and ideas presented, evaluate the value and relevance of material and where appropriate apply that knowledge to their future study. A programme of related reading was also recommended to students. Assignments submitted, however, indicated that sometimes, even where students had attended a lecture on the topic, they did not appear to have acquired or used the relevant information in a critical way.
on lectures and directed reading tasks
Students were asked to keep a lecture journal in which they recorded both lecture notes and subsequent reflections upon the content and its relevance. As a inter-lecture activity students met in "critical friends" groups, read each others files and gave each other feedback on the effectiveness of their note making and associated commentary. Both feedback and responses to the feedback were recorded and the quality of this dialogue became part of a profile of formally assessed coursework.
This approach to peer assessment proved to be very time consuming and placed greater demands upon the students. There were logistical problems for the groups in trying to arrange meetings at mutually convenient times and places and many groups found it hard to resolve interpersonal difficulties. The groups ability to deal with differing levels of commitment between group members was a particularly influential issue. Students were divided in their opinions about the value of the peer feedback. Some thrived on the challenge of collaboration and found genuine mutual support, encouragement and a high quality of critical feedback within their groups; they felt they had both given and received help which led to improvements in both their understanding of the field of study and their general abilities as learners. Others remained frustrated by the process which they viewed either as slowing their own pace or limiting access to more expert assistance which was seen as remaining in the domain of the lecturer.
3) Peer assessment of written assignments
Students were required formally to mark and moderate examples of each other's assignments. The first assignment of the year was given a strict limit of 1,000 words and all students were required to submit their scripts in type with personal information limited to a removable frontispiece. This allowed for the preservation of confidentiality. One seminar session was then given over to the marking task.
Students were placed in groups of four. Each group was responsible for the marking of four scripts. Working initially in isolation students had ten minutes to mark a script, followed by five minutes in which they recorded marks and comments against a set of published assessment criteria. The sequence of events was strictly timed and orchestrated by the lecturer, so that after one hour each student had marked all four assignments. The group then had half an hour in which to compare their responses and agree upon a final mark for each script and write detailed comments for the author on the strengths and weaknesses of the assignment in relation to each of the given criteria.
Initially the students were highly resistant to the suggestion that they were to mark each other's work. They felt it:
After considerable persuasion one group of 75 students agreed to take part in the exercise (although they insisted that all the work be second marked by the lecturer). In practical terms the activity was carried out with very few problems. The students treated the exercise extremely seriously, they were easily kept on schedule and by the end of the one and a half hours, every paper had been marked four times and at least a full side of A4 comments prepared as feedback for each.
- was inappropriate: the lecturer is paid to do it;
- would lead to invalid assessments: the lecturer was qualified to make judgements while they were not;
- was an intrusion on their academic privacy, no-one other than the lecturer usually saw their work;
On completion of the exercise the student response was quite remarkable: Every participant was enthusiastic, positive and convinced that it had been an excellent learning opportunity. Many described it as the most significant learning event in their tertiary experience (this particular group were post graduates and therefore had at least three years prior tertiary experience). Out of the 75 students only three had ever read another student's completed assignment, even though many more claimed to have collaborated in assignment writing. They were amazed by the differences they had found between work in terms of quality and approach. They quite shocked to find that each marker interpreted the criteria differently but they felt they greatly clarified their own understanding of the assignment brief and criteria for marking, through their need to reach group agreement on grades and comments. The students universally acclaimed the exercise as having enlightened them about their own work. In making judgements about other student work they had clearly referred constantly to comparisons with their own and this seemed to bring both strengths and weaknesses into sharper focus. Every student was able to identify at least one aspect of their writing which they could improve immediately and without additional help; simply seeing the way others had met criteria well provided a variety of role models to follow or adapt with confidence. The students generally found the task exhausting and expressed both sympathy and admiration for the academic staff who regularly undertook such work.
The papers were later remarked by lecturing staff (without reference to the student grades and comments). The consistency of marking between students and staff was astonishing. There were no instances of significant disagreement, and only one case where the final grade differed by more that one grade.
Evaluation of the effectiveness of the changes
Learning is a complex affair and formal evaluation of methods trialed was not conducted at a sufficiently rigorous level to allow for legitimate claims to be made for achievement of superior learning. External examiners of the course did, however, comment at the end of the year on the exceptionally high level of critical thinking and debate evident amongst both staff and students. Students were generally able to articulate well about their learning; provide evidence of development in their knowledge, skills and attitudes; and identify future directions for their learning. Student evaluations of the course indicated that students believed that they had work harder than in previous courses and that they had achieved a depth of understanding beyond the ability to recall given information. In addition they acknowledged the complexity of their field of study and the need to continue a programme of self and professional develpoment beyond the course. Although student responses to different strategies varied, the overall response was favourable and even where students did not enjoy active participation in peer assessment they were able to acknowledge many advantages and benefits.
In the short term the teaching workload was not significantly reduced since the introduction of the programme of peer assessment itself was labour intensive (staff meetings, development of procedures, confirming student assessments by second marking etc). There was certainly, however, the promise of lowering lecturer workloads where confidence in the procedures could be established as a standard part of the course.
Lecturers generally found that they became more committed to peer assessment through the trial period. In part this was due to the evidence of good learning outcomes amongst many students, and the potential they saw for relieving some of the assessment workload. It also related to a strengthened belief in the goals of student self direction, self evaluation and the aim of encouraging learners who could work independently and who made a lifelong commitment to continuing development.
|Please cite as: Sparrow, H. (1997). Peer assessment as a tool for learning. In Pospisil, R. and Willcoxson, L. (Eds), Learning Through Teaching, p307-311. Proceedings of the 6th Annual Teaching Learning Forum, Murdoch University, February 1997. Perth: Murdoch University. http://lsn.curtin.edu.au/tlf/tlf1997/sparrow.html|
[ TL Forum 1997 Proceedings Contents ]
[ TL Forums Index ]
HTML: Roger Atkinson, Teaching and Learning Centre, Murdoch University [email@example.com]
This URL: http://lsn.curtin.edu.au/tlf/tlf1997/sparrow.html
Last revision: 10 Apr 2002. © Murdoch University
Previous URL 18 Jan 1997 to 10 Apr 2002 http://cleo.murdoch.edu.au/asu/pubs/tlf/tlf97/spar307.html