200 Human Resources Management for Public and Nonprofi t Organizations
(Schmidt, 1988). Hunter (1986) found the following evidence in regard to
cognitive ability tests:General cognitive ability predicts performance ratings in all lines of
work, though validity is higher for complex jobs than for simple jobs.
General cognitive ability predicts training success at a uniformly
high level for all jobs.
Data on job knowledge show that cognitive ability determines how
much and how quickly a person learns.
Cognitive ability predicts the ability to react in innovative ways to
situations in which knowledge does not specify exactly what to do.Measurement Terms and Concepts
Employers need to be familiar with a number of psychometric concepts
that pertain to employment testing.Reliability This concept of reliability is concerned with the consistency of
measurement. An exam ’ s reliability can be determined through a number
of different procedures. Test - retest reliability occurs when individuals taking
the test score about the same on the test in each administration. If a test is
reliable, there should be consistency between two sets of scores for the test
taken by the same person at different times. Split - half reliability is derived by
correlating one part of the exam with another part of it. If the exam is mea-
suring an aptitude reliably, it should do so throughout the exam. Odd - even
reliability is when a score is computed for all the even - numbered exam items
and then correlated with a score derived from the odd - numbered items.
Internal consistency measure reliability is when each exam item is correlated with
every other exam item. Equivalent - forms reliability is when different forms of
an exam have been constructed. Each version of the exam is administered
to the participants, and the two sets of exam scores are correlated.(^) Validity Is the test or selection instrument measuring what it is intended
to measure? Validity is the most important characteristic of measures used
in personnel selection. Why use a particular test or procedure if it is not
predicting or evaluating correctly the most qualifi ed candidates? Valid-
ity and reliability are often confused. Reliability is necessary for a test to
be considered valid, but it cannot stand alone; just because a test gives
consistent results does not mean that it is measuring what it is intended to
measure.