CONJUGAL RELATIONSHIPS BETWEEN ASSESSMENT AND
RESEARCH
While educational assessment tries to model that which is latent (i.e. cannot
be seen, felt, heard, or even sensed) , in an individual, using the results of a
confrontational interaction between the trait or ability involved and
appropriate tasks; educational research generally tries to model a
population's behaviour, which is also latent, based on what is observed to
hold for a representative sample. Hence both assessment and research share
a lot in common as inferential processes, which try to find out the truth about
individual behaviour, in the case of assessment; and about mostly population
behaviour, in the case of research. Three types of variance that influence
these processes: systematic desirable variance from the trait or ability under
assessment or the independent variable in research which must be
maximized; systematic error variance emanating from other extraneous
sources in both assessment and research, which must be controlled; and the
ever-present random error variance which must be minimized in other to
attained reliable results.
Key words: Assessment, research, measurement, error, systematic desirable variance,
systematics error variance, random error variance.
H. Johnson Nenty, PhD
CONJUGAL RELATIONSHIPS BETWEEN ASSESSMENT AND RESEARCH
While educational assessment tries to model that which is latent (i.e. cannot be seen, felt, heard, or even sensed) , in an individual, using the results of a confrontational interaction between the trait or ability involved and appropriate tasks; educational research generally tries to model a population's behaviour, which is also latent, based on what is observed to hold for a representative sample. Hence both assessment and research share a lot in common as inferential processes, which try to find out the truth about individual behaviour, in the case of assessment; and about mostly population behaviour, in the case of research. Three types of variance that influence these processes: systematic desirable variance from the trait or ability under assessment or the independent variable in research which must be maximized; systematic error variance emanating from other extraneous sources in both assessment and research, which must be controlled; and the ever-present random error variance which must be minimized in other to attained reliable results. Key words: Assessment, research, measurement, error, systematic desirable variance, systematics error variance, random error variance.