Rethinking Evaluation of Student Learning
Patricia Cranton
University of New Brunswick
Abstract
This paper presents a critical examination
of strategies for the evaluation of student learning. A perspective on
evaluation is proposed which takes into account Habermas’s three kinds of
knowledge: instrumental, practical or communicative, and emancipatory. The
research paradigms corresponding to each kind of knowledge are used as a
starting point for thinking about the types of evaluation of learning strategies
that might be appropriate. It is suggested that objectively-scored techniques
are suited to the assessment of instrumental knowledge, interpretive techniques
to the evaluation of communicative knowledge, and subjective self-evaluation to
the understanding of emancipatory knowledge.
Introduction
T |
here may be no other everyday issue with which faculty struggle as
much as evaluation of student learning. We want to be fair to students. We try
to be objective. We want grades to accurately reflect what students know. We
understand that students compete for grades and are motivated by marks. We
realize that students rely on grades as an entry to further studies, jobs, or
career advancement. Our classes are often large, and we feel forced to use
quick-to-score multiple choice tests even while we doubt their validity.
There are no clear-cut solutions to these
issues. Evaluation of learning is a complex process that takes place within
many different contexts and is conducted by individuals with widely varying
values and philosophies of teaching. However, I believe many of us have fallen
into certain mindsets regarding evaluation, limiting our ability to see
alternatives and fresh perspectives.
In this paper, I suggest we rethink and
critically examine some of our taken-for- granted assumptions about evaluation
of student learning. I propose a perspective on evaluation that takes into
account the nature of the knowledge we want our students to attain. Although
this sounds quite ordinary, and like something we already do, I show that such
a rethinking leads to reflection on several of our assimilated norms about
evaluation. Perhaps objective is not always better than subjective. Perhaps
there are even no truly objective ways to evaluate learning. Perhaps teachers
are not always the best judge of what students have learned.
First, I use Habermas’s (1971) three kinds
of knowledge as a starting point for this discussion. Habermas’s classic work
on human interests and how knowledge about them is acquired has spawned a rich
literature in education. I suspect his ideas can also guide us in reexamining
such practical issues as how to evaluate learning. Second, I draw a parallel
between evaluating learning and educational research methodologies as a way to
spark new thoughts about evaluation. Third, I contemplate how these
perspectives might translate into strategies we can use in higher education.
This leads me to three kinds of evaluation: objectively-scored, interpretive,
and subjective or self-evaluation.
Three Domains of Knowledge
H |
abermas (1971) describes three basic human interests, each of
which leads people to acquire a different kind of knowledge. One of his theses
is that instrumental knowledge has become all-pervasive in modern society,
thereby enticing us to look at everything through this lens. I think this may
be true of how we view and value student learning. But before I discuss this
possibility, I will briefly review the three domains of knowledge.
Our technical interests lead us to instrumental knowledge.
Technical interests come from the need to control and manipulate the external
environment so as to obtain, for example, shelter, food, transportation, and
fuel. Instrumental knowledge is based on invariant causeand-effect principles.
It allows us to predict observable events in the world. The knowledge is
objective and empirically derived. Instrumental knowledge falls into the
philosophical realm of positivism. Habermas’s distinction between invariant
cause-and-effect relationships that are true in all situations and those that
vary with social factors and changing conditions is important to note here.
Because of the invariant laws of gravity, for example, a stone thrown in the
air will always fall to the earth. Human learning, thinking, development, and
emotions cannot be predicted in the same way, however much we would like it to
be so.
Our practical interests lead us to
practical knowledge or communicative knowledge, the latter term being one
employed by Mezirow (1997, for example) in his work on adult learning. Mutual
understanding of individual interests and needs is required for people to live
together in a society and to coordinate social actions to satisfy those
interests and needs. We have to live together and work together. In order to do
so, we need to understand each other on both a simple personal level and on a
larger social and political level. Practical knowledge is acquired through
language, through communication with each other. It is a knowledge of the norms
that underlie the society we live in, whether this be interpersonal
relationships, groups, communities, organizations, cultures, nations, or the
global society. Rather than truth in terms of invariant principles, as exist in
instrumental knowledge, validity is determined by consensus within a group and
a sense of “rightness” or morality. What is agreed-upon knowledge in one
culture may not be valid in another culture. The philosophical foundation,
therefore, for communicative knowledge lies in hermeneutics. Our justice
systems, social systems, and political systems are based on communicative
knowledge. Values related to family, work, individualism, and spiritualism are
communicative in nature.
Our emancipatory interests are reflected
in our desire to grow and develop, to obtain self-knowledge, and to experience
freedom and relational autonomy. Emancipatory knowledge is acquired through
critical reflection and critical self- reflection (Mezirow, 1991).
Philosophically, the underpinnings lie in critical theory. We are constrained
by uncritically assimilated norms, beliefs, and values. Social norms and social
systems become reified, unquestioned, and unquestionable. When we absorb these
norms and systems, we are oppressed or constrained in that we are not aware of
alternatives. For example, as a young girl growing up in the 1950s in rural
Western Canada, I assimilated the social norm that girls could become wives,
mothers, teachers, or nurses. No other alternatives existed in the community
nor in my mind. I was constrained by that norm. Our basic human drive for
growth can lead us to critically
question such systems and thereby become free from the constraint of not
knowing there are alternative perspectives.
In higher education, the acquisition of
each of these kinds of knowledge are valid and accepted goals, though there is
a tendency to value instrumental over communicative and emancipatory knowledge
(for example, see Cranton, 1998). This is a remnant of the Age of Enlightenment
when instrumental rationality was seen to be the solution to the world’s ills.
A reverence for objective science remains with us still. It is in the sciences,
of course, where we find the acquisition of instrumental knowledge to be a
primary goal of higher education. The trades and technologies also focus mainly
on instrumental knowledge in their programs. Communicative knowledge is the
purview of the social sciences, humanities, and arts. As well, the service
industries and helping professions seek to foster communicative learning to a
large extent. Unfortunately, it is often the case that in these disciplines,
researchers, theorists, and teachers strive to force knowledge into the
instrumental domain in an attempt to give it greater value (for example, see
Kincheloe, 1991). In the social sciences (note the assumption underlying the
terminology we use), the search for invariant relationships still dominates a
good portion of academic research and writing. Emancipatory knowledge
essentially crosses disciplines. Most faculty advocate critical thinking, for
example, which is the basis for questioning assumptions and beliefs. And, at
least in our rhetoric, we aspire to student self-determination and autonomy.
Self-determination includes both being reflective about cultural contexts and
traditions and being articulate about our own values. Autonomy or empowerment
is achieved through self-reflection on the subjective conditions of knowledge
(Kant’s critique of knowledge) and reflection that frees the individual from
the hidden constraints inherent in social structures (Marx’s critique of
ideology). Self-determination and self-reflection at least touch all
disciplines and may be seen as the primary goal in some programs.
Research Methodologies
T |
he impetus for this essay came from comments I received from a
reviewer of a chapter on transformative learning. He or she asked for empirical
evidence (in the narrower empirical-analytical sense rather than the broader
experiential sense) regarding transformative or emancipatory learning. This led
me to reflect on how we think of evidence of student learning. It is quite
clear that different research methodologies are required for obtaining
instrumental, communicative, and emancipatory knowledge. It may seem to be an
aside to bring research methodologies into this discussion when rethinking evaluation
of student learning is the focus. However, higher education faculty are in tune
with the way research strategies vary according to the question being
addressed, and I see the acquisition of knowledge through research as parallel
to the acquisition of knowledge we expect of our students. A brief examination
of research paradigms may stimulate our thinking about evaluating learning.
Each scientific methodology is a valid
means for knowing within its domain of knowledge. Instrumental knowledge is
acquired through the empirical-analytical or natural sciences (Ewert, 1991).
Quantitative measurement, experimental design, and the traditional scientific
method for addressing research questions are employed. Research questions are
generally of the forms, “Does A cause B?” “Is there a significant difference
between X and Y?” “How do A and B vary together (or correlate)?” and “How
frequently or where does X occur?” The goal of empirical- analytical research
is to discover invariant relationships. We strive for correct or true answers
to questions. Social factors do not influence the outcome. The researcher is
objective; her opinions do not have an impact on results.
Communicative or practical knowledge is
obtained through the hermeneutic or interpre- tive sciences (Ewert, 1991).
Qualitative data are used to interpret and understand intersubjective meaning.
Through interviews, observations, immersion in a context or situation, written
materials, and conversations with others, researchers strive to understand human
and social processes (for example, see Creswell, 1998). Research questions are
generally of the form, “What is the nature of X?” “How do people understand Y?”
and “How do individuals learn about X?” The goal is not to establish causality
but to inquire into the meaning of things in the world. The researcher is an
interpreter rather than an objective observer.
Emancipatory knowledge is obtained through the critical sciences
(Ewert, 1991). Unlike
the empirical and interpretive sciences that describe the world as
it is, the critical sciences critique the world with a goal of knowing how it
should be. The purpose is enlightenment and emancipation. Self-reflection and
self-development are explicit aims for the researcher working in this paradigm.
The researcher therefore works with participants who are seen as coresearchers
in a transformative process (for example, see Merrian and Simpson, 1995).
Research questions take the form, “What is happening here?” “How did we come to
this state?” and “How can we open up alternative perspectives and ways of
acting?” Participants not only seek an understanding of a situation, but they
also jointly make decisions about possible courses of action.
Parallels Between Research Paradigms and
Evaluation Strategies
T |
here is, of course, a fundamental difference between conducting
research and evaluating student learning. Regardless of what is being learned,
learning is influenced by social and psychological factors. Therefore it never
falls entirely into the instrumental domain. If we understand this, though, I
believe we can still use the parallels between research and evaluation to spark
ideas about evaluation.
When our goal is to evaluate student learning in the instrumental
domain, the nature of the knowledge allows us to employ quantitative, objective
strategies, as it does when we engage in empirical-analytical research. When
there is one correct answer to a problem or clearly correct procedures for
completing a task, we can judge whether or not students have acquired the knowledge
by their ability to produce the correct answers or demonstrate the correct
procedures. We could argue that though students may learn in different ways or
at varying speeds all that matters in the end is the quantifiable and
observable learning outcome. This is a commonly-held view in many higher
education programs. Psychological factors such as learning styles and social
factors such as classroom dynamics are things to be dealt with in our teaching
methods, not our evaluation strategies. I am not sure I would actually advocate
this position, but it is one that follows logically from focusing on
instrumental learning outcomes. It is worth considering, even as we also
critically question it.
When our goal is to evaluate learning in
the communicative domain, we are looking at quite a different picture. Using
the hermeneutic or interpretive research paradigms as a model, we are still
interested in describing what happens in the objective world, but we are now
concerned with an interpretive understanding of what we see. There are no
objectively verifiable truths. Validity is determined by consensus within a
social group of informed persons. In this domain, student learning should then
be evaluated by methods which allow freedom of expression and interpretation
(essays, oral presentations, role plays, creative and artistic productions),
interpersonal interactions (discussion, conversation, interviews), and
flexibility in form and content. Attempts to quantify and objectify learning in
this domain deny its very essence. This is a key point and one that follows
naturally from thinking about the parallel research paradigm. In evaluating
learning, we have traditionally striven for objectivity. Yet, the knowledge itself is such that it
needs to be interpreted so as to make meaning of it. As faculty, we are
informed experts in our field, as is a researcher interpreting qualitative
data. We can validly judge the meaning of students’ work.
When it is our goal to evaluate learning
in the emancipatory domain, we are interested in how self-reflection and
self-development are taking place. In the critical sciences research paradigm,
participants (those we are striving to understand) are fully engaged as
coresearchers. It is the person who undertakes self-reflection and self-
development who can say that he or she is doing so. The researcher helps
participants to understand meaning, interprets actions, challenges assumptions,
and asks good questions. But participants must be involved for the research to
take place as the process is wholly subjective. If we take apply these notions
to evaluating students’ emancipatory learning, we are led to student
self-evaluation. This is not to say that students can assign themselves marks
or grades without dialogue and guidance, any more than research participants
can address the research questions without the researcher, but they must be
actively involved in the evaluation process in a way that promotes
self-determination. If they are not, the goal of emancipatory learning is
defeated.
Implications for Selecting Evaluation
Strategies
T |
raditional instructional design procedures include clarifying the
learning objectives along with their domain and level of learning in order to
best select testing methods. In one way, what I am suggesting here is similar,
except in traditional instructional design the goal is always to find
observable (quantifiable, empirical) indicators of learning. I recall some of
the texts from the late 1960s and early 1970s that advised us to list
observable indicators of students’ motives, attitudes, emotions, and values.
This led to such bizarre statements as, “The student will voluntarily attend
art gallery showings eight times out of ten and stay for at least one hour per
visit,” as an indication of student interest in art. The word ‘subjective’ was
a derogatory term. Although instructional designers had the right idea when
they advocated matching evaluation strategies to the nature of learning, they
then unfortunately forced everything into the instrumental mould. Under the
influence of behaviourism, those were also the days when writers in education
advised the use of task and procedural analyses of learning objectives,
competency-based modules, standardized curriculum packages, and standardized
test scores.
When instrumental knowledge is being
evaluated objectively-scored assessment techniques are generally appropriate.
By objectively-scored techniques, I mean those for which two people using a
scoring key would come up with the same result. Multiple-choice, truefalse,
matching, and some short answer formats are included. I deliberately say
objectivelyscored rather than objective since the selection and formulation of
the questions themselves involves subjective judgement. Some objectively-scored
tests are very subjective indeed, reflecting the teacher’s perception of what
is important in the course and his or her biases in terms of wording, weights
on questions, and so forth. Unfortunately, when we quantify things such as
student learning, we tend to give the numbers, the results, more power and
precision than they should have simply because they are numbers and we have
learned to value numbers.
The evaluation of communicative learning
must be explicitly and openly interpretive, if we follow through from our earlier
analogy of research paradigms. This is not a label found in the literature on
evaluation of learning. I use the term interpretive evaluation techniques to
include the familiar written essay, term paper, or report. Case studies and
problem solving exercises are relevant in some disciplines. Student
presentations, discussions, debates, and other forms of oral communication also
fall into this category. The traditional thesis defence is an example of an
oral evaluation strategy upon which we place great value in higher education.
In some disciplines, student performance (drama, music, sports) or products
created by students (art, architecture, engineering) are interpreted in order
to evaluate learning.
This may seem to be nothing new. Do we not
already use these methods regularly? I am suggesting that we use such
techniques in a different way, or perhaps that the philosophy underlying their
use be rethought. Many faculty using, for example, essays or term papers try to
make the scoring as objective as possible. They count up points, subtract
points for spelling errors, follow predetermined criteria, and still feel
vaguely uneasy about the scoring. What about the original essay that meets few
of the predetermined criteria? What about the student who has the courage to
challenge existing viewpoints? What about the student who has made enormous
strides in her thinking but has still listed all the main points? In nearly 25
years of academic life, I do not think I have met a faculty member who enjoys
marking essays. This could be because we are forcing an interpretive procedure
into an objective model. Don’t art and film critics enjoy judging the quality
of the work they review? Perhaps this is due to the freedom they have to be
openly interpretive.
Good interpretive evaluations are
trustworthy and credible. In accepting this form of evaluation, we rely on the
expertise, professionalism, and credibility of the teacher or any other judges
who might be involved. We rely on negotiation, agreement, and consensus among individuals.
Talk of “bias” is irrelevant. Our inclinations, beliefs, and values become a
part of the evaluation process. Evaluations are based on our interpretations
rather than detached from them. The trustworthiness of the evaluation is
founded on the professionalism, experience, knowledge, authentic, and ethical
behaviour of the educator.
To value interpretation in evaluation of
learning challenges our assumptions. The behaviourists and instructional
designers, and the pervasiveness of instrumental knowledge have been strong
influences in education. To interpret does not mean to be random, unfair, or
discriminatory against those students who have perspectives different from our
own. In interpretive evaluation procedures, there should be agreement between
faculty and students that the nature of the learning is communicative. There
should be agreement as to what is being evaluated. The person doing the
evaluation should be a subject expert with established and respected
credentials. Evaluators must be ethical, caring, responsible, and open-minded.
Most of us would agree that these conditions are not onerous. Most of us would
also agree that interpretive evaluation is valid and credible in many other
aspects of life. When we cannot accept it in the evaluation of communicative
learning, we need to step back and question our basic assumptions.
The evaluation of emancipatory learning
may require an even bigger stretch. In writing about transformative
(emancipatory) learning, Mezirow (1991, pp. 219-220) makes a strong and
challenging statement: “dogmatic insistence that learning outcomes be specified
in advance of the educational experience in terms of observable changes in
behavior or ‘competencies’ that are used as benchmarks against which to measure
learning gains will result in a reductive distortion and serve merely as a
device of indoctrination.” As we saw earlier, based on the parallel between
research paradigms and evaluation models, student self-assessment needs to be
central to the evaluation of emancipatory learning. Evaluation is subjective,
of the self. If our goal is truly emancipatory, it clearly makes no sense to
not allow student involvement in one of the most powerful aspects of the
learning experience – judging its quality. Empowerment on a short leash is not
empowerment at all. We cannot say, “Our goal is self-direction,
self-determination, and self-reflection, but I will judge how well
you did.” Even though it may be well-meaning and most often based on our
uncritically assimilated assumptions about the role of teacher, I would argue
that this stance is hypocritical.
How can self-evaluation be incorporated
into the higher education setting where students are often motivated by grades
and competitive in attaining them? Several points can be made in response to
this question. Perhaps we need to question the systems that have led to grade
motivation and competition. But that is a topic for another time. For our
purposes here, it is important to realize that self-evaluation is critical when
the learning goal is emancipatory. I believe that subjective self-evaluation
should be encouraged in all learning, but there are many occasions where this
does not work, given the nature of higher education institutions. So, first, we
need to be clear about when and how we are promoting emancipatory learning and
promote self-evaluation in relation to those goals.
Good self-evaluation focuses on
self-reflection, openness to alternative points of view, an increased awareness
of underlying assumptions, participation in reflective discourse, and changes
in long-established patterns of expectations and behaviours. Students know when
they have learned in these ways. We may not.
But even so, it is not so simple as
saying, “Student, evaluate thyself.” Students need to learn how to engage in
self-evaluation. They may need help clarifying goals and how their learning
relates to those goals. Students can help each other in this process, and
faculty should also try to meet individually and privately with students to
discuss their self-evaluation. Examples of how to judge emancipatory learning
can be given. Students should be encouraged to validate their perceptions with
another person—the teacher, a peer, a professional in the field. Students have
well-established habits of mind about the teacher’s role in evaluating their
learning. It may be as difficult for them to participate in evaluation as it is
for many faculty to invite it. Students will say, “You’re the teacher – you
tell me,” reflecting a whole web of assumptions about what teachers and
students should do. It is our responsibility to challenge those assumptions.
Critical Questions for Reflection
I |
have suggested that an
understanding of Habermas’s three kinds of knowledge and the research paradigms
that are used to obtain those kinds of knowledge leads us to three quite
distinct approaches to evaluation of students’ acquisition of knowledge.
Objectively-scored techniques can be used to assess students’ instrumental
knowledge. Interpretive strategies can be used to evaluate students’
communicative knowledge. Subjective self-evaluation is appropriate for
understanding students’ emancipatory learning. Rather than formulating a
conclusion, I would prefer to end this discussion with some questions that we
may all take away and ponder as we go about our practice.
Is it possible to be truly objective in
the evaluation of student learning which is, by definition, a psychological and
social process? Even when the knowledge attained is instrumental in nature,
should we take into account how students learn? How can we account for, in
striving for objectivity, the personality, preferences, and values of the
person who designs the evaluation method?
In the interpretive model of evaluation,
can we still compare students? How do we translate evaluation results into the
numerical or letter grades acquired by most institutions? Can we be sure that
all faculty interpreting student work are indeed professional, responsible,
ethical, and caring? What arguments can we present to our colleagues to
persuade them that it is valid to interpret students’ work? Are interpretive
methods appropriate only in the arts and social sciences, or are they also
relevant at advanced levels in the sciences?
How can subjective self-evaluation be incorporated into institutional
requirements? Can we expect honest and valid subjective self-evaluation from
students when they are also competing for grades? How can we best prepare
students to engage in self-evaluation? Do undergraduate students have the
maturity and self- awareness to judge their progress toward emancipatory
knowledge? If we advocate self-determination and critical reflection in higher
education, how can we deny students the opportunity of self-evaluation?
References
Cranton, P. (1998). No
One Way: Teaching and Learning in Higher Education. Toronto, Canada: Wall
& Emerson.
Creswell, J.W. (1998). Qualitative
Inquiry and Research Design. Thousand Oaks CA: Sage Publications.
Ewert, G. (1991).
Habermas and education: A comprehensive overview of the influence of Habermas
in educational literature. Review of Educational Research, 61, 345-378.
Habermas, J. (1971). Knowledge and Human Interests. Boston:
Beacon Press.
Kincheloe, J. (1991). Teachers
as Researchers: Qualitative Inquiry as a Path to Empowerment. London:
Falmer Press.
Merriam, S.B., &
Simpson, E.W. (1995). A Guide to Research for Educators and Trainers of
Adults. Malabar, FL: Krieger Publishing Company.
Mezirow, J. (1991).
Transformative dimensions of adult learning. San Francisco:
Jossey-Bass.
Mezirow, J. (1997). Transformative learning: Theory to practice.
In P. Cranton (Ed.), Transformative learning in action: Insights from
practice. New Directions for Adult and Continuing Education, no. 74. San
Francisco: Jossey-Bass.