英语论文网

Now,as we know from our daily experiences,inferences and surmises can be
erroneous.In order to constrain our inferences,we develop procedures for gathering information that we may use as a basis for the conclusions we wish to draw.This involves two stages.First,we build a model of the domain or criterion;we call thisthe test construct.This will be subject to ongoing revision as our understanding ofthe domain,and oflearning,changes over time.The source of the modelling may ofcourse be the curriculum,which itself constitutes a model,a view of the target
domain,its structure and principal divisions and characteristics,and assessments mayaccordingly draw on curriculum constructs to enable us to interpret the data of learnerperformance.But because constructs also embody social,educational and politicalvalues,and are never neutral,they are constantly subject to contestation and socialand political influence,as we shall see.Finally,we use the construct,our understanding of what it is we would like to know,to develop a procedure for principled observation of performance under knownconditions.We call this the test or the assessment.If we have set the thing upproperly,we are then in the position to draw inferences on the basis of theseobservations either about the probable character of performance under non-testconditions,or about the candidate’s standing in relation to a domain of knowledge andabilities in which we are interested.
In summary,then,we make a distinction between the criterion(the target of testinferences)and the test(the evidence from which we draw our inferences).Andbecause our inferences are necessarily indirect,they are mediated through testconstructs,that is,our modelling of the criterion in terms of its essential features orcharacteristics.5
THE ASSESSMENT ARGUMENT
Central to assessment,then,is the chain of reasoning from the observations to theclaims we wish to make about students–Mislevy calls this the‘assessment argument’(Figure 2).This is needed to establish the relevance of assessment data and its valueas evidence.
Observations
(Assessment
data)
a
via
Assessment
argument
(relevance of
data;its value
as evidence)
aClaims about
students
(inferences)
Figure 2 The assessment argument
According to Mislevy,
An assessment is a machine for reasoning about what students know,can do,or
have accomplished,based on a handful of things they say,do,or make in
particular settings.Mislevy,Steinberg and Almond(2003),p.4.
Let’s take each part of this definition in turn.
An assessment is a machine for
reasoning
ASSESSMENT ARGUMENT6
The term‘machine’here is perhaps puzzling:the reference is not to anything physicalor technological,but is used metaphorically,meaning a‘mechanism’or‘systematicprocedure’.The key element in assessment then is‘reasoning’–the adequacy of ourrepresentation of the criterion in the test construct,and the relationship of theassessment procedures to that construct.about what students know,can do orhave accomplishedCLAIMS
We begin with what we would like to be able to assert about thelearner,in terms ofwhat students‘know,can do,or have accomplished.’For this,Mislevy usesthe terms‘claims’.What claims would we like to be able to make about a learner on the basis
of our handful of observations?What are our certificates attesting to?This requiresus to think through what we want to be able to say.Here,senior secondarycertificates are in a