Notes on:
Byrne, C. (1975) 'Assessment systems and student
learning'. In Baggaley, P. et al (eds) Aspects
of Educational Technology VIII.
Pitman Press: Bath: 197--205
Dave Harris
There are problems in assembling a number of
assessment results into an overall
system. Assessment can be formative or
summative, with the former tending to be
informal, although there is now a trend to
formality. Summative assessment is seen
as the statement of what has been learned, and
is useful for employers and others.
Assembled assessment results can take the form
of:
An index, A conflated score, often
weighted. The rationale for the weights
can be a problem. It might be connected
to external criteria, but this is difficult:
for example job requirements differ. The
index can conceal variations.
A profile. There is no
conflation. Independent sections of
courses are assessed. Scores can be
tested [for reliability] by correlations
between the tests of performances on two
sections. The separate dimensions of the
profile can be supplied from a 'psychometric
approach'.
Problems with conflation arise because if the
components are separate enough to warrant
separate tests, how can they be related?
All is well if the components can be seen to
build to mastery, but there is still a problem
of disentangling the effects on the overall
score of earlier knowledge that is required
[which students bring to the course].
Only estimates are available.
The use of continual assessment
throughout the course often simply assumes
that there is this building towards
mastery. It is sometimes also assumed
that final examinations lack validity that is
they are scored unreliably, so that additional
items are needed. Continuous assessment
can remove uncertainty for students [and also
cool them out better!], But there are still
some problems:
Items are valid only if there is an agreement
on what is being tested. Examinations
need a link to suitable formative
assessment. Adding items does not always
increase the validity overall. Student
efforts are harnessed best through formative
assessment. Students like assessment to
give them some direction and control over this
study—giving equal weight to each assignment
might motivate students, but this can also
make no sense in terms of the logic of
building towards some summative outcome.
It is necessary to link up the formative and
summative in order to provide more credibility
with the former [but you do get rehearsal
effects]
more education
studies
|
|