Grading should reflect how well students
have achieved the intended learning outcomes for the
subject or programme. The principle is the same as
Constructing Assessment Criteria
for particular assessment tasks, and as such it should
be criterion-referenced
not norm-referenced (CRA not NRA).
In developing grading criteria, a framework is
required to enable teachers to describe and conceptualise
the very best (A+) and the least acceptable (D)
performances (see qualitative
assessment). Counting "marks" is
usually not the best way to go (see quantitative
assessment). The institution may develop an
overall framework for grading, but individual departments
will need to develop content specific criteria of
what they mean by "A-ness", "B-ness"
etc. In traditional honours degrees, this is well
worked out already for the different classes of
degrees. The SOLO taxonomy can be useful for deriving
such qualitative categories.
SOLO taxonomy
SOLO stands for Structure of Observed Learning
Outcomes. (Biggs and Collis, 1982) As students learn,
the outcomes of their learning display stages of
increasing structural complexity. The taxonomy distinguishes
between five levels of understanding: prestructural,
unistructural, multistructural, relational, and
extended abstract. These five levels are hierarchical,
with extended abstract representing the highest
level of understanding.
Level |
One
will be able to
|
Prestructural |
| |
No understanding. One simply misses the point. |
|
|
Unistructural |
| |
Possess knowledge of a single fact about something, e.g. one begins to be
able to name things.
Examples: name, label, identify, recognise, execute simple procedures |
|
|
Multistructural |
| |
Recite multiple but isolated facts about something.
Examples: list, describe, match, outline |
|
|
Relational |
| |
One begins to see relationships between things and be able to form a holistic
picture of the issue.
Examples: apply, explain, analyse, discuss, compare, relate |
|
|
Extended abstract |
| |
At this level, original thinking is evident and one begins to generate new
ideas that go beyond what is already known to him/her.
Examples: theorise, criticise, create, generalise, hypothesis, design |
|
|
Source: Modified from Biggs, J (2003). Teaching for Quality Learning at University, 2nd Edition, Society for Research into Higher Education & Open University Press.
The above is only a generic model, it
needs adapting to suit particular tasks in different
disciplines. There are many examples of such uses
of the SOLO
taxonomy on the internet.
Arriving at an overall grade from a number of assessment
tasks
Even though the initial assessment has been graded
qualitatively, combining performances in several assessment
tasks is easily done by converting grades into numbers,
combine and then re-convert back to grades.
Arriving at an overall grade can also be handled
qualitatively by using profiling.
Profiling
Profiling is one of the ways to handle the problem
of weighting and combining assessment results. When
all tasks are of equal importance and each is graded
qualitatively, the final grade can be derived from
the most typical or modal response. If a student is
mostly working at B level, a grade B is assigned.
In case of uneven profile, the highest level the student
has achieved could be taken as the final grade on
the grounds that the student has demonstrated this
level of performance in at least one task. A conversion
can also be used: A= maximum performance on all tasks;
B= maximum on two tasks, very good on remaining ones;
C= one maximum, two very good, rest pass, and so on.
Some assessments require the students to demonstrate
different levels of performance in different tasks
and hence required a weighted profile. For instance,
Task A might require a high level of understanding,
Task B declarative knowledge of selected topics, and
Task C correct terminology only. All have to be passed
at the specified level. Weighting in this case is
not an arbitrary juggling of numbers but a profile
determined by the structure of the curriculum objectives.
For further information, see Biggs (2003, pp197-207).
|
