Developing a typology for learner self-assessment practices

Abstract

The use of student self-assessment is becoming increasingly common in university courses. Self-assessment is not a unitary idea: it means different things to different practitioners and the practices associated with each conception lead to quite different outcomes. Most accounts of self-assessment in the literature have tended to obscure these differences and treat each example as if it were pursuing the same goals as every other. The paper distinguishes self-assessment from related ideas such as self-testing, self-rating and reflective questioning. It points to key issues which need to be considered in designing self-assessment activities and proposes a threefold categorisation which highlights the links between self-assessment and the knowledge interests which are being pursued.

Full text

Paper published in Research and Development in Higher Educ ation, 18, 130-135, 1995.
Developing a typology for learner self-assessment practices
David Boud, University of Technology and
Angela Brew, University of Portsmouth and University of Sydney
Abstract
The use of student self-assessment is becoming increasingly common in university
courses. Self-assessment is not a unitary idea: it means different things to different
practitioners and the practices associated with each conception lead to quite different
outcomes. Most accounts of self-assessment in the literature have tended to obscure
these differences and treat each example as if it were pursuing the same goals as every
other. The paper distinguishes self-assessment from related ideas such as self-testing,
self-rating and reflective questioning. It points to key issues which need to be considered
in designing self-assessment activities and proposes a three-fold categorisation which
highlights the links between self-assessment and the knowledge interests which are
being pursued.
Introduction
The use of learner self-assessment is being driven both by educational considerations and
by the need for staff time to be most effectively used. As self-assessment has proliferated, it
has become apparent that many different notions are being subsumed under this general
heading. Indeed, it is increasingly difficult to know what is meant when it is reported that
self-assessment is being used. The aim of this paper is to explore this diversity, identify what
might properly be included within the notion of learner self-assessment and provide a
framework for discussing different approaches. This paper draws on material contained in
the book Enhancing Learning through Self-assessment (Boud, 1995).
What is self-assessment?
The defining characteristic of self-assessment has been expressed as:
the involvement of students in identifying standards and/or criteria to apply to their
work and making judgements about the extent to which they have met these criteria and
standards. (Boud 1991, p. 5)
This reflects the two elements of any assessment process: the identification of standards
with specific criteria related to these and the making of judgements based on them.
Self-assessment is used as both a verb and as a noun: a process, as well as an activity with a
distinct identity. It is a practice in which to engage as well as a goal to which to aspire.
Without reference to the context it is often hard to discern which of these two uses is
intended. It is important to make the distinction because self-assessment is a goal of higher
education which may be pursued through course designs which do not involve self-
assessment exercises as such. However, as many courses in higher education have been

designed in ways which inhibit the development of self-assessment skills, a useful first step
is the introduction of explicit self-assessment activities.
It is important to identify what the definition implies and does not imply as there are many
misconceptions about self-assessment. [The term self-evaluation is also used, particularly in
North America—see, for example, MacGregor 1994 who uses self-evaluation to mean self-
assessment without any component of grading]. There are a number of common practices
which are sometimes referred to as self-assessment, but which are sufficiently different to
warrant separate consideration and the use of alternative descriptors. These are self-
testing, self-rating and the use of reflective questions.
Self-testing involves students checking their performance against provided test items. These
may include the common text book examples, the answers to which may be found at the
back of the book, or items specially designed by teachers. A correct or desired answer is
normally specified by a teacher or course writer. Responses which do not conform to this
are not regarded as legitimate. Either the ‘answer’ alone may be provided or test items may
be linked with explanations (eg. model answers) or study materials. Devices to avoid
students looking at an answer before working out the problem are sometimes used. These
often involve withholding ‘correct’ answers until items have been completed. However, it is
sometimes difficult to imagine why teachers go to such great lengths to inhibit learners
devising their own way of approaching learning tasks.
Self-rating on a provided scale is a variation on this theme. Self-rating approaches normally
do not involve answers which are right or wrong. Each student’s responses are as legitimate
as any other’s, although in some situations, a particular range of responses might be
regarded as more socially desirable than others. The purpose is to have students reflect on
their present state of knowledge or achievement. An example of this is the core competency
ratings which Knowles’ (1990) uses to help learners diagnose their learning needs. Other
examples are found in management courses (eg. the Harvard MBA) in which students use a
wide variety of personality, learning style, personal preference instruments to develop a
profile of their own strengths and weaknesses. Self-rating is often used in the development
of student profiles in areas such as the recording of personal transferable skills.
A common feature of both self-testing and self-rating is that students are not normally
expected to actively engage with or question the standards and criteria which are used.
Criteria may be explicit (in the case of model answers with detailed marking instructions) or
implicit (in the case of back of the book answers), but it is not expected that these will be
modified by the student or that alternative ones will be proposed.
A third type of activity which is akin to self-assessment is the use of reflective questions. It is
common in distance learning materials to include in-text questions to prompt learners to
reflect on what they have been reading. While some of these may take the form of self-
testing with answers provided, often such questions are included to help students engage
with the material and relate it to their own situation. It is an open question whether such in-
text questions help or hinder learning. They may, for example, disrupt the acquisition of
knowledge from text by drawing attention away from the normal devices which authors use
to signal intent and importance. Reflective questions can also be used in a class setting

either as a means of checking understanding or to promote student reflection which goes
beyond the material at hand.
Reflective questions are included here as they can appear similar to self-tests or self-ratings.
However, they may more appropriately be located in a typology of reflective approaches in
courses where they could take their place alongside such strategies as journal keeping,
debriefing, and mind mapping (Boud & Knights, forthcoming). There are clearly connections
between reflection and self-assessment. Both involve focusing on learning and experience,
but self-assessment is usually concerned with the making of judgements about specific
aspects of achievement, often in ways which are publicly defensible (eg. to teachers),
whereas reflection tends to be a more exploratory activity which might occur at any stage of
learning and may not lead to a directly expressible outcome. All self-assessment involves
reflection, but not all reflection leads to a self-assessment. As we shall see below, there are
some forms of self-assessment which demand critical reflection.
Why is a typology needed?
It is necessary to discriminate between different types and approaches to self-assessment to
avoid the confusion which exists when a single term covers a multitude of sometimes
incompatible practices. A typology is needed both for purposes of course and subject design
and for research and evaluation on self-assessment. If the key variations can be identified, it
may be possible to ascertain the advantages and disadvantages of each and where they
might most appropriately be used. If self-assessment has greater potential for enhancing
learning than as a substitute for teacher-conducted assessments, as has been argued
elsewhere (Boud, forthcoming), it is necessary to be clear about what aspects of self-
assessment contribute to what kinds of learning.
The quality of research studies on self-assessment leaves much to be desired. Among the
studies examined by Boud and Falchikov (1989) there were a wide variety of elementary
methodological weaknesses. Very few, indeed, were free of technical flaws, eg. the scales
used were not specified, or teachers used different criteria to students. There was
insufficient reporting of the contexts of the studies, eg. it was often not clear whether the
results of self-assessments were used for formal grading purposes or even what it was that
was being assessed. There was relatively little cross-referencing of publications: it appears
that the errors in one study were not spotted and corrected in later ones. Their analysis of
such flaws led Boud and Falchikov to propose standards for the reporting of empirical
studies of self-assessment.
However, it is necessary to go beyond the methodological imperatives which they discussed.
At the most basic level it is vital to be clear what practice is being reported. Is it self-
assessment of the kind defined above, or one of the many forms of self-testing? Many
empirical studies make the assumption that the most important concern of self-assessment
is whether or not student marks coincide with those of teachers. This is relatively easy to
determine if both parties are using common standards and criteria, but much more
problematic when students engage with criteria which vary from those of teachers. The
conceptual difficulties of undertaking empirical research in such circumstances means that
there is little systematic evidence on more sophisticated versions of self-assessment. A

framework for classifying self-assessment practices is a prerequisite for further research in
this area.
Dimensions which might be considered
There are a number of ways in which self-assessment practices can be differentiated. We
would expect variations on all of the following dimensions:
• Subject matter domain and specific learning goals. Self-assessment is used in many
different subject areas from the highly technical or skills based to the discursive or
expressive.
• Domains of learning such as knowledge and understanding, skills and capability, and
competency and learning outcomes. External influences on higher education are leading
to greater emphasis being placed on the assessment of general skills and attributes and
on the use of competency frameworks and outcome-oriented measures.
• Students’ prior experience of tasks of the kind being assessed, level of study, or students’
stage of learning. A minimum level of knowledge is required in any area before the
notion of self-assessment becomes meaningful and the kinds of self-assessment task are
likely to vary as students develop more sophisticated conceptions of the area they are
studying.
• Cognitive complexity of the task and the kinds of knowledge schema required. Research
on cognition suggests that learners need to develop schemas to make sense of what they
are learning and how it fits with other ideas before they can proceed further.
• Student self-confidence or self-efficacy. While these influence how students respond to
any given self-assessment activity, they also influence what self-assessment demands it is
realistic to make on students.
• Students’ experience of self-assessment processes, in particular of the specific type being
used. Evidence suggests that students’ skill in self-assessment develops through practice,
so the form chosen is likely to be influenced by what has gone before.
• Degree of involvement of other parties. This might include feedback from peers, input
from staff or the use of members of a profession.
• Whether the outcomes of self-assessment are to be used solely for learning or whether
they are associated with grading and the formal accreditation of learning. The need to
impress external audiences may overwhelm the making of an assessment which is true to
the learner.
A typology of self-assessment would need to take account of and, possibly, to integrate
these dimensions. We wish in the remainder of this paper to propose a classification which
we have found productive in our analysis of self-assessment practices and which we believe
will throw light on some major differences which can be observed. It is particularly useful in
highlighting what different self-assessment practices are trying to achieve. It links to a
broader theoretical framework about knowledge and it provides a basis for making
judgements about the effectiveness of self-assessment practices. The classification concerns
the knowledge interests pursued by particular tasks.

Categorisation of self-assessment according to knowledge interests
Different kinds of knowledge interests are served by different kinds of learning and
assessment tasks. This follows Habermas’ idea that knowledge is always shaped by the
needs and desires of human beings and that different kinds of knowledge give expression to
different ‘knowledge constitutive interests’ (Habermas 1987). Examples of self-assessment
can be seen to lie within one or other of the three knowledge constitutive interests of
Habermas: the technical interest, the communicative interest and the emancipatory
interest. These knowledge interests are present whether what is being assessed is students’
knowledge and understanding, academic, personal or transferable skills or whether they are
assessing their own competency or learning outcomes (Brew, forthcoming).
Technical interest: Habermas’ project was to show that knowledge does not exist
independently of us. Scientific knowledge is built up because humans have a desire to
exercise control over their world. They attempt to exercise this control by seeing the world
as separate objects which can be observed and measured and about which predictions can
be made. Such knowledge has, what Habermas terms, a technical knowledge interest,
‘interest in technical control over objectified processes’ (1987: 309). This interest is not
confined, however, to the sciences. It can be seen to be served whenever objectified
statements are present. As far as self-assessment is concerned, we say that a technical
knowledge interest is being served where students are engaged in checking their
knowledge, skills, understanding or competency against a set of more or less objective
statements. Such statements are, like scientific ‘facts’ presumed to exist more or less
independently of each other and to represent discrete aspects of the world.
Assessment of knowledge in the technical interest is, by and large, a question of checking
what facts, ideas, etc. or how much information has been understood. Similarly, in this
conception, skills are viewed as objectified skills. In many cases, they tend to be seen more
or less independently of each other. The context in which they are developed is not
considered particularly relevant in assessing them. Indeed, the notion of transferability
includes the idea of being able to use a skill in a variety of contexts. Similarly, competencies
in the technical domain are viewed as separate objectified statements of expected
performance. Assessment is again a question of checking whether they have each been
achieved. Indeed, in all of these cases, the simplest form of self-assessment is a set of
statements, the students’ task being to assess how far they have achieved them. Other self-
assessment activities include students checking answers against a model answer sheet, or
with respect to a prepared set of criteria or competency statements on scales which have
been provided to them.
Communicative interest: Sometimes knowledge is more a question of interpretive
understanding and objectivity is problematic. Habermas views this kind of knowledge as
being constructed in a process of mutual negotiation and communication. Such knowledge
has what he terms a practical or communicative knowledge interest which he says aims at
action-oriented mutual understanding (1987: 310). In considering the scope of self-
assessment a practical knowledge interest is being served when the self-assessment
involves elements of communication and interpretation, perhaps a negotiation of criteria, or
discussion of the relationship of one element of the assessment to another.

In this domain, a number of different and sometimes competing conceptions of reality may
come together. In the process of understanding, students develop their personal meaning.
They make socially constructed knowledge their own. From a communicative perspective,
skills are viewed as part of a process of negotiation; the outcomes of some kind of dialogue
or conversation. Transferable skills are likely to be integrated into academic study. They are
viewed as part of a composite of skills within a broad capability in each person. Indeed, skills
are more likely to be seen as integral to the process of acquiring knowledge. In this domain,
competency is viewed as a professionally defined construct (cf. the holistic notion of
competencies described by Hager, Gonczi & Athanasou, 1994).
Examples of self-assessment within the communicative interest include students discussing
what constitutes a ‘good’ answer and researching or negotiating criteria which they then
apply to their work. For example, students in a class test situation outline an answer to a
question. One student describes her answer and others add ideas to form a collective view
of what is a good answer. Students then mark their own answers in relation to this. They
then go on to the next question and so on to the end of the test (Pastol 1993).
Emancipatory Interest: But people do not only either build up objective knowledge or
engage in a process of interpretive understanding. They also reflect on the processes in
which they are engaged. Knowledge therefore, Habermas suggests, includes a meta-level
analysis where reflection is part of the process of building knowledge. This third domain of
knowledge pursues what Habermas terms an emancipatory knowledge interest. It is
concerned with the pursuit of reflection as such, ie. self-reflection is part of the process. It is
emancipatory in the sense that the meta-level analysis has the capacity to bring into
consciousness the very ways in which that knowledge is constructed and therefore to go
beyond it. Self-assessment where students radically change perspective by reflecting
critically on the determinants of their work exemplifies this third knowledge interest.
Students not only construct meaning for themselves, they also come to a critical
understanding of the way in which that is done.
In the emancipatory perspective, meta-level skills including critical reflection may be
developed through a change in teaching towards an emphasis on student autonomy and
responsibility or through other strategies where there is a questioning of assumptions.
Where competencies are concerned, criticism of the competencies themselves is a
characteristic of this mode. Competency statements are viewed as attempts to make
relatively stable statements of learning outcomes which are to be questioned through the
practice of self-assessment. In the emancipatory conception, skills, knowledge and
understanding, competencies and learning outcomes are often difficult to separate.
In the emancipatory domain discussions which take place are not about whether the
student has met a set of negotiated criteria. Here the criteria and students’ understanding
of them develops as they proceed. Students do not simply accept the standards or a set of
given competencies. They engage in a critique of them and this informs their judgements
about their work. Some of the emancipatory examples use reflective writing for learners to
focus on a particular learning experience and to form judgements about it.

An example of self-assessment in the emancipatory domain in the subject area of computer
science can be found in Edwards (1989). Students perform tasks, set up their own
competency standards and assess themselves. They then justify these to their tutor and
may change their assessment or the criteria in the light of these discussions. Other examples
of self-assessment in the emancipatory area in other disciplines can be found in Boud
(1992), Cowan (1988), Kramp & Humphreys (1993) and Moore & Hunter (1993).
Conclusion
This paper has explored some of the dimensions of self-assessment which must be
considered in the development of a typology which will aid research and practice. We have
proposed a classification of self-assessment practices based on the different knowledge
interests which they serve. This provides the basis for discussions about the key factors to
be taken into account in the selection and design of such practices and goes some way
towards supplying the foundation for a typology. Further conceptual work is needed to
clarify the significance of some of the dimensions discussed here and more empirical studies
of self-assessment are required to identify those features which are most salient for student
learning.
References
Boud, D. (1992). The use of self-assessment schedules in negotiated learning. Studies in
Higher Education. 17, 2, 185-200.
Boud, D. (1995). Enhancing Learning through Self Assessment. London: Kogan Page.
Boud, D. & Falchikov, N. (1989). Quantitative studies of student self-assessment in higher
education: a critical analysis of findings. Higher Education. 18, 5, 529-549.
Boud, D. & Knights, S. (forthcoming). Course design for reflective practice. In Gould, N. &
Taylor, I. (Eds.). Reflective Learning for Social Work: Research, Theory and Practice.
Aldershot, Hants.: Ashgate Publishing.
Brew, A. (forthcoming). Self-assessment in different domains. In Boud, D. Enhancing
Learning through Self Assessment. London: Kogan Page.
Cowan, J. (1988). Struggling with student self assessment. In Boud, D. (Ed.). Developing
Student Autonomy in Learning. Second Edition. London: Kogan Page, 192-210.
Edwards, R. M. (1989). An experiment in student self-assessment. British Journal of
Educational Technology. 20, 1, 5-10.
Habermas, J. (1987). Knowledge and Human Interests. Trans. Shapiro, J. London: Polity
Press. First published in German 1968 by Suhrkamp Verlag.
Hager, P., Gonczi, A. & Athanasou, J. (1994). General issues about assessment of
competence. Assessment and Evaluation in Higher Education. 19, 1, 3-16.
Knowles, M. S. (1990). The Adult Learner: A Neglected Species. Fourth Edition. Houston,
Texas: Gulf Publishing.
Kramp, M. K. & Humphreys, W. L. (1993). Narrative, self-assessment and the reflective
learner. College Teaching. 41, 3, 83-88.
MacGregor, J. (1993). Learning self-evaluation: challenges for students. In MacGregor, J.
(Ed.). Student Self-Evaluation: Fostering Reflective Learning. New Directions for Teaching
and Learning No. 56. San Francisco: Jossey-Bass, 35-46.

Moore, W. S. & Hunter, S. (1993). Beyond ‘mildly interesting facts’: student self-evaluations
and outcomes assessment. In MacGregor, J. (Ed.). Student Self-Evaluation: Fostering
Reflective Learning. New Directions for Teaching and Learning No. 56. San Francisco:
Jossey-Bass, 65-82.
Pastol, G. (1993). Turning a test into a learning situation. Methomix: Occasional Publication
of the Teaching Methods Unit. Rondebosch: University of Cape Town. 15, 1, 3-4