Edusfarm, revista d’educació superior en Farmàcia. Núm. 1, 2007.
PEER ASSESSMENT AS AN APPROACH TO JUDGE GROUP WORK: DOES
IT WORK?
Vicente Rodilla
Universidad Cardenal Herrera CEU, Dep. Fisiología, Farmacología y Toxicología,
Área de Toxicología
Avda. Seminario s/n, 46113 Moncada, Valencia
Rebut: maig de 2006. Acceptat: desembre de 2006
ABSTRACT
Peer assessment strategies have been around for quite a while and the benefits derived from
their implementation have been described and demonstrated. Peer assessment increases student
motivation, builds up on self confidence and stimulates critical thinking and is a particularly
useful strategy to assess the contribution of each student to group work. It is however a type of
assessment which can be difficult to implement if the students do not support it. In this paper we
describe and discuss the relative lack of success we have had when we have tried to use self and
peer assessment to evaluate group work in Spain, whereas a very similar strategy worked
extremely well for us in the United Kingdom. We also discuss the reasons which may explain
this apparent contradiction.
KEY WORDS: assessment, peer assessment, group-based assessment, evaluation.
RESUMEN
La estrategias de evaluación que realizan los estudiantes sobre sus propios compañeros (peer
assessment) se conocen desde hace unos años y se han descrito y demostrado los beneficios que
de su utilización se derivan. Estos métodos evaluativos incrementan la motivación de los
estudiantes, refuerzan su autoestima y estimulan el razonamiento crítico. Es además una
estrategia muy útil para determinar la contribución que realiza cada estudiante cuando se trabaja
en grupo. Es sin embargo un tipo de evaluación que puede resultar difícil de implementar si no
recibe el apoyo de los estudiantes. En este artículo describimos y discutimos el relativo poco
éxito que hemos tenido cuando hemos tratado de utilizar autoevaluación y peer assessment para
evaluar el trabajo en grupo en España, mientras que una estrategia similar nos funcionaba
1
Rodilla, V.. Edusfarm, revista d’educació superior en Farmàcia. Núm. 1. 2007
_____________________________________________________________________________________
extremadamente bien en el Reino Unido. Discutimos también las razones que podrían explicar
esta aparente contradicicón.
PALABRAS CLAVE: evaluación, peer assessment, evaluación de grupos.
INTRODUCTION
In recent years, social and emotional gains attract as much interest as cognitive gains and to this
aim, cooperative and group learning, and peer assessment, play a very important role in a
number of educational establishments, from kindergarten to secondary school (Topping, 2005).
This is also true for Universities where a relative big weight is being placed on learning and
developing transferable skills, without compromising the acquisition of knowledge by the
students.
It could be argued that because of the inherent gregarious nature of the human being, teaching
and learning in groups is probably a more natural approach than the individual approach until
now so prevalent in formal education. It has been shown that cooperative group work improves
in the students the acquisition of information and its retention as well as social and
communication skills and self-confidence (Johnson et al., 1998). Therefore group work can be a
very important teaching strategy in any course, in as much as it helps to develop teamwork
skills without sacrificing the acquisition of knowledge. An important question that arises when
group or collaborative work is implemented is how to assess the work the students have carried
out with fairness. The lecturer can evaluate the final outcome of collaborative work, but it
would be impossible for the lecturer to measure and assess the contribution of any individual
student within the group. Who is then best placed to carry out a fair and objective assessment of
the individual performance of each student within the group? It is certainly not the lecturer,
particularly given the fact that much of the work has been undertaken away from the classroom
and away from his/her supervision. This assessment is better carried out by the members of the
group: only the students know, and can evaluate therefore what has been the contribution of
each one of the members of the group to the task assigned.
Assessment is an intrinsic part of any teaching method. It has been stated that much of what
happens in university studies is driven by assessment, and it has been estimated that teachers
can spend between one third to half of their teaching time dealing with matters which are related
directly or indirectly to assessment (Schaffer, 1991). However, effective assessment is more
often than not difficult to implement adequately to make it a part of the teaching process by
2
Rodilla, V.. Edusfarm, revista d’educació superior en Farmàcia. Núm. 1. 2007
_____________________________________________________________________________________
which student learning is enhanced and reinforced. Sally Brown (1999) maintains that the
methods used to assess students in Universities and Colleges around the globe are extremely
limited. In fact, she has estimated that approximately 80% of total assessments are traditional
exams, essays and reports (Brown, 1999). Most of these more traditional forms of assessment
are generally teacher-centred assessments and therefore they limit the opportunity for the
students to fully understand educational objectives and furthermore, they limit the opportunity
to develop critical evaluation skills by the students. Other less conventional assessment
approaches would be self-assessment, peer assessment and group based or computer based
assessments. Amongst the different approaches and methods used to assess students, one that
has been successfully described by some authors and which is becoming widely used in a
number of Universities around the world is peer assessment (Brown & Knight, 1994; Topping,
1998; Billington, 1997; McLaughlin and Simpson, 2004; Kaufman et al., 2000; Elliott and
Higgins, 2005). Peer assessment provides feedback opportunities for the students as they are
assessing other students. Some papers promote peer assessment usage and several authors have
tested its validity and reliability in a number of applications such as student presentations, group
work, projects, reviews or essays handed out to the class, etc. (Billington, 1997; Topping, 1998;
Topping, 2005).
An intra peer assessment in which the students assess the performance of other students with
whom they have been working is particularly useful in as much as it allows the lecturers to
explore and obtain important information about group dynamics. If properly performed, intra
peer assessment allows the lecturer to detect individual student contribution to the overall
product. It may be used to detect leaders in each group, and perhaps more importantly it is
invaluable to detect those students who are not pulling their weight or are simple hitchhikers
(those who take a free ride at the expense of their more industrious colleagues) (Kaufman et al.,
2000). Since one would tend to think that the students would be in favour of a fair and equitable
system of assessment for group work and sometimes are concerned that the group work does not
reflect their own individual effort, peer and in some cases self- assessment can be the answer. In
fact, we would argue that for assessing group work, peer assessment is about the best way for
the lecturer to get a fair idea of the contribution each student has made to the overall project.
For several years, we successfully used self and peer assessment to judge the contribution of
individual students to group work in a Biomedical Sciences course in the UK. In this paper we
describe and discuss the difficulties we have experienced in implementing peer assessment to
judge group work with Spanish students. It was never our intention to carry out a controlled
experiment by comparing two different populations of students to peer assessment. What we
3
Rodilla, V.. Edusfarm, revista d’educació superior en Farmàcia. Núm. 1. 2007
_____________________________________________________________________________________
present here is rather the results which have appeared after several years of using peer
assessment to evaluate the group work of our students both in the United Kingdom and in Spain.
METHODOLOGY
As stated above, we developed a peer assessment strategy for group work assessment which was
implemented in three successive academic years (from 1997-98 to 1999-00) at the School of
Applied Sciences at the Robert Gordon University (Aberdeen, Scotland) and an assessment pro
forma was designed based on
Stefani and Tariq (1996) (Figure
1). This assessment pro forma
was translated into Spanish and
used for self and peer
assessment when group work
was involved at the Faculty of
Experimental and Health
Sciences at the Universidad
Cardenal Herrera CEU
(Valencia, Spain) (Figure 2). In
both universities, the reasons as
to why peer assessment was
necessary and the importance of
peer assessment to the overall
assessment of the group work
was explained to the students.
Using this assessment form the
students had to grade
themselves and each member of
their group on several aspects of
the group work, which were
Figure 1. Self and peer assessment pro forma which
we used at the Robert Gordon University, Aberdeen, attendance at group meetings, their
to assess student contribution to group work. contribution with ideas and
suggestions, efficiency in carrying out the tasks assigned by the group, acceptance of their fair
share of work, contribution to the finished project (poster, presentation, report, etc) and their
overall contribution to the project. In each category, the student marked themselves and their
peers on a scale from 0 to 6. (Figure 1). At the Robert Gordon University peer assessment for
4
Rodilla, V.. Edusfarm, revista d’educació superior en Farmàcia. Núm. 1. 2007
_____________________________________________________________________________________
group work amounted to 10% of the mark for the group assignment. The lecturer (or lecturers)
grading of the group work contributed a 50-60% and the remaining 30-40% was obtained either
by the students assessing the group assignment or in some cases from the assessment grades
given to the group by all the other groups. At the Universidad Cardenal Herrera during the years
2000-01 to 2002-03, when a group exercise was carried out and the attendance to laboratory
practicals was approximately a 30% of the final mark. Self and peer-assessment varied from 20
to 40% of the total mark for group assignment. The mark given by the lecturer to the group
work had a weight of 60 to 80%.
Figure 2. A used example of a pro forma used at the Universidad Cardenal Herrera CEU to
evaluate student contribution group work. The surnames of the students have been deleted.
RESULTS AND DISCUSSION
When self and peer assessment was used with our Biomedical Science students at the UK, the
results were highly satisfactory and allowed us to evaluate group work with fairness. While
some students expressed their doubts and reservations about assessing their peers the first time
this system was used, particularly if they were considering giving a low mark to any colleagues,
those doubts were not an issue in subsequent years, although they had to be reassured that the
grades given to any student would be treated confidentially.
5
Rodilla, V.. Edusfarm, revista d’educació superior en Farmàcia. Núm. 1. 2007
_____________________________________________________________________________________
On one occasion while teaching on a module on Human Anatomy and Physiology the group
work did not involve peer assessment and the comments the students made during the
evaluation of the course were very clear (Table 1). The students were clearly against a single
mark given equally to each individual within the group and wanted to assess their own
contribution.
For several years we have tried to put into practice this peer assessment approach with our
Pharmacy and Veterinary students at the Universidad Cardenal Herrera but with limited success.
The first year we tried to use it, everybody in the class awarded themselves and their peers in
the group either a 5 or a 6, which invalidated the assessment, because all students in the group
had the same marks. Kaufman and co-workers (2000) have described similar situations, with
students agreeing to give one another identical ratings and they see this fact as evidence that the
teams are working correctly and everyone participates evenly in the group work. However in
their experiment, this happened in 5-10% of the teams, which can be acceptable. We have had a
similar situation is Scotland. With such a low incidence, one cannot discard that this is a
Table 1. Comments made by students when peer assessment was not used to evaluate their
non-standard assessment (NSA).
NSA should be able to be marked, at least in part, on an individual basis
Change the NSA (We must assess our project group)
NSA should be on individual basis, not group.
Students should be marked on their own input/knowledge to the NSA.
Group assessments are very unfair, and not a true reflection on each individual
coincidence because the group is working well and everyone pulls their weight, or whether the
students have reached an agreement indeed. In further years we have tried different
modifications to make peer assessment work, either by explaining in much more detail the
importance of fair assessment and what we wanted to get from the students marks, or by adding
further rules to the marking system. For instance, if in the group there are n students the same
mark could only be used n-2 times in each category, but this strategy complicates the
assessment unnecessarily.
In both cases and with a number of groups the students had made a prior pact, so that everybody
in the group had the same overall mark. When this happens, the whole exercise is pointless,
because we (the lecturer) fail to find out, and therefore assess, what the contribution of each
student has been to the group project.
6
Rodilla, V.. Edusfarm, revista d’educació superior en Farmàcia. Núm. 1. 2007
_____________________________________________________________________________________
Average peer assessment 6
5
Nevertheless on some
4
occasion such as the
3 Veterinary (2003-04) or
2 Pharmacy students (2002-
1 03) self and peer
assessment was correctly
0
0 1 2 3 4 5 6 used and with a certain
Self-assessment degree of success. For the
Figure 3. Scatter plot showing the relation between self and
average peer assessment marks for 2002-03 academic year sake of the discussion we
for Toxicology students in the Pharmacy degree.
present here the results
for the Pharmacy course taught in 2002- 03. In Figure 3 we show the correlation between the
self-assessment and the average peer assessment marks on a group exercise. There is a
statistically significant positive correlation between the marks the students award themselves
and the average awarded to each student by their group colleagues (ρ=0.754, Spearman rank
test; p < 0.05). Half the students (n= 22) awarded themselves a slightly higher mark that that
their colleagues awarded, whereas the opposite, colleagues awarding a higher mark than
themselves happened with 17 students (38.6%). With 5 students (11.4%) the grade awarded by
colleagues or by themselves was identical. One student got a particularly low mark by their
peers, but he was aware of his own contribution and also awarded himself a low mark. These
results are similar to those described by other authors (Kaufman et al., 2000). One team was not
included in these calculations since it was considered to be a dysfunctional team: three members
of the team failed to submit their ratings. Even on these occasions in which self and peer
assessment seems to work, the students remain unconvinced about their utility since they did not
like the idea of grading themselves and less, grading their peers.
During the year 2004-05, and given the, at least from our point of view, unsatisfactory results
obtained in the previous years, a group exercise was put into practice which did not involve peer
assessment. Each group of students prepared an 10-15 minute oral presentation on the toxic
effects of a selected drug (or a family of compounds) to be given in class to their peers and they
also submitted a written report. The lecturer assessed the presentation and the report, grading the
group work and allocated each student in the group the same mark irrespective of their
contribution. We expected to hear some comments about the fairness/unfairness of this system,
and how single student input was not taken into consideration, but this was not the case. In
contrast to their British counterparts (see comments on table 1), the Universidad Cardenal
7
Rodilla, V.. Edusfarm, revista d’educació superior en Farmàcia. Núm. 1. 2007
_____________________________________________________________________________________
Herrera CEU students seemed to be quite happy sharing the marks irrespectively of their
contribution. They certainly did not make a single comment about being unfair or wanting to
assess themselves individually.
According to the literature (Kaufman et al, 2000; Felder and Brent, 1994), and in our own
experience, the most common complain the lecturer receives when the possibility of group work
within the class is mentioned, is about hitchhikers, but when the students know that their mark
will effectively lower the mark of that particular student but not lower the marks of all the other
working members of the team they cease to complain about the unfairness of group work.
However we have not got this type of complain with our Pharmacy or Veterinary students in
Spain. In most cases in fact, the students resent having to downgrade a colleague because
his/her contribution to group work is not acceptable and they tend to cover for that particular
student.
The marks that were obtained from the peer assessment were substantially increased from a
10% of the total mark in Scotland to a 40 % in Spain. This was done because the students in
Aberdeen always complained that the proportion of marks allocated to peer assessment was too
small. Similar comments have been reported by others who also allocate a 10% to peer marking
(Elliott and Higgins, 2005).
The question we ask ourselves is “why did self and peer assessment worked so well in The
School of Applied Sciences in Aberdeen and is not well received in Spain?” Felder and Brent
(1994) have reported that lecturers who have attempted to install cooperative learning in their
courses “frequently encounter resistance and sometimes open hostility from the students”. This
is probably something that has happened to us since our Veterinary and Pharmacy students are
not as familiar with these types of strategies as their Scottish counterparts. In fact non standard
assessments, whether group or individually based, and using a variety of different approaches
(posters, seminars, oral presentations, written work, etc) were a common occurrence for most
modules (subjects) in a degree or even in the Diploma courses. There are also differences in our
students and our Universities, which we have not taken into account. In Scotland, most students
either live on-campus or nearby; at Universidad Cardenal Herrera-CEU, many students live off
campus, some needing to travel an hour or more to come to college. Spanish timetable is also
very different from the timetable in other European countries. Scottish students would be
required to be at college during the day from 9:00 h to attend lectures, laboratory practicals,
meetings, etc., but their college engagements would be over by 17:30 h at the latest. The
number of subjects (or modules) in which the student registers is much more regulated by the
8
Rodilla, V.. Edusfarm, revista d’educació superior en Farmàcia. Núm. 1. 2007
_____________________________________________________________________________________
Faculty and the University and as a result, all (or at least most) students enrolled on a course
follow either the same or a very similar timetable. The students at Universidad Cardenal
Herrera-CEU have a broader timetable with lecturers, practicals and other activities distributed
(almost at random) throughout the day. To begin there is the lunch break which can occupy two
to two and a half hours in the middle of the day. Additionally, there are practically no
restrictions on the subjects the student chooses to study in a particular year. Thus, a student may
begin the day at 8:30 with the first lecture and have the last at 19:00 h. In between there are
many free hours (one here, two there, another elsewhere) depending on the number of subjects
that particular student is registered. Time availability to do group collaborative work in Spain
can be therefore much more restricted. All these constrictions can make meetings between
students outside lectures, if not impossible at least, very difficult.
There are also educational differences in the students when it comes to assess their peers.
Spanish students often question the validity of these exercises, being reluctant to grade
themselves and even more reluctant to grade their peers. They tend to see downgrading a
student who has not contributed to the group work as a telltale activity; worse still they think the
lecturer should not find out. On occasions the group has been discovered to be covering for
students who do not pull their weight to the team work. However, the attitude in Scotland also
reported by others (Elliot and Higgins, 2005; Felder and Brent, 1994) is very different, with the
students having no problems down-grading individuals who were not contributing what they
should to the group effort.
Given the demonstrated benefits of cooperative learning coupled with self and peer assessment,
by means of which the students are better motivated, tend to show higher academic
achievements, achieve better high-level reasoning and critical thinking skills, normally through
working in groups with their peers gain a deeper understanding of the learned material, develop
their teamwork skills, etc ( Felder and Brent, 1994; Elliot and Higgins, 2005, Stefani and Tariq,
1996; Topping 1998, 2005; Johnson et al., 1998; Hanrahan and Isaacs, 2001; Gatfield, 1999) we
must continue to try to welcome this type of learning amongst our students.
Perhaps where collaborative learning and peer assessment are difficult to implement, as it is our
case, we (the lectures) should try to make the students our allies by involving them more
actively in the design of the assessment forms and the computations for the overall marks for
student and group work, so that they can perceive it as a fair system of awarding marks. A
number of authors (see Elliot and Higgins, 2005) have demonstrated that self and peer
assessment strategies are well received by the students when they are actively involved in them.
9
Rodilla, V.. Edusfarm, revista d’educació superior en Farmàcia. Núm. 1. 2007
_____________________________________________________________________________________
CONCLUSIONS
While it is generally agreed that a diversity of assessment approaches are needed in order to
assess students fairly, not all of these strategies are useful in all settings. For instance, peer
assessment is not well received by Spanish students when compared to their British
counterparts. In our experience, Spanish students are extremely reluctant to participate in self or
peer assessment exercises. They tend to see peer assessment as a telltale activity short of being
moles or defectors. They also consider it to be an interference in their own world. British
students are also far more competitive amongst themselves than Spanish, to the extent that in
some cases they do not even share lecture notes. These very different attitudes could be a very
important factor which may help to explain why some, successfully proven, teaching strategies
do not seem to work with certain groups of students. Perhaps there is a philosophical and
theoretical framework which explains and directs the context in which peer assessment takes
place and it is this framework and its connotations that needs to be discovered and exploited to
make a fuller use of peer assessment.
REFERENCES
Billington, H.L. (1997). Poster presentations and peer assessment: novel forms of evaluation and
assessment.
Brown, S. (1999). Promoting Effective learning through assessment. The Robert Gordon University,
Aberdeen.
Brown, S. and Knight, P. (1994). Assessing learners in higher education. Kogan Page, London.
Gatfield, T. (1999) Examining students satisfaction with group projects and peer assessment.
Assessment and Evaluation in Higher Education. 24: 375-377.
Hanrahan S.J. Isaacs, G. 2001. Assessing self and peer assessment: the students’ viess. Higher
education Research & Development. 20: 53-70
Elliott, N., Higgins, A. (2005). Self and peer assessment – does it make a difference to student group
work? Nurse Education in Practice, 5: 40-48.
Felder, R.M., Brent, R. (1994). Cooperative learning in technical courses. ERIC Document
Reproduction Service Report ED 377038. (http://www.ncsu.edu/felder-
public/Papers/Coopreport.html).
Kaufman, D.B., Felder, R.M. , Fuller, H. (2000). Accounting for individual effort in cooperative
learning teams. J Engr Education 89: 133-140.
Johnson, D.W., Johnson, R.T., and Smith, K.A. (1998). Active learning: cooperation in the college
classroom. Edina, Interaction book Co. Minnesota.
10
Rodilla, V.. Edusfarm, revista d’educació superior en Farmàcia. Núm. 1. 2007
_____________________________________________________________________________________
McLaughlin, P., Simpson, N. (2004). Peer assessment in first year university: how the students feel.
Studies in Educational Evaluation, 30: 135-149.
Schaffer, W.D. (1991). Essential assessment skills in professional education of teachers. Education
Measurement Issues and Practice, 10: 3-12.
Stefani, L.AJ. , Tariq, V.N. (1996). Running group projects for firs-year undergraduate students. J
Biol Education 30: 36-44
Topping, K. (1998). Peer assessment between students in colleges and universities. Rev. Educ. Res.
68: 249-276.
Topping, K. (2005). Trends in peer learning. Educational Psychol, 25: 631-645
11
