Salvia and Ysseldyke ( 1985) define assessment as, "the process of collecting
data for the purpose of (1) specifying and verifying problems, and (2) making
decisions about students" (p. 5 ). Five primary types of decisions that can be
made from assessments have been identified. These are decisions about referral,
screening, classification, instructional planning, and pupil progress evaluation.
For each of these decisions, academic, behavioral, or physical problems may be
the targets of assessment.

The interaction of decision type and problem area results in 15 different
possible kinds of assessment. Although school psychologists may be involved in
any of these, the largest proportion of time spent in assessment by school psy-
chologists usually concerns decisions around classification and instructional
planning ( Goldwasser & Meyers, 1983). The types of decisions to be made
through assessment, however, are not usually specified by the referral source.
Teachers refer because they are having a problem with a student and would like
assistance. Once referred, a process is begun to understand why the child is
having the problems that have lead to the referral and to respond to questions
being asked by the referring teacher. Although the referring source does not
differentiate the type of decision to be made, each of these decisions should
require a differential assessment process ( Salvia & Ysseldyke, 1985; Ysseldyke
& Mirkin, 1982). It makes little sense to be using instruments designed to make
classification decisions to prescribe remediation strategies.

Unfortunately, this has not been the case. Ysseldyke, Algozzine, Regan, and
McGue ( 1981), in a computer simulation, found that a cross-section of educa-
tional personnel including school psychologists, regular education teachers, spe-
cial education teachers, administrators, and pupil personnel support staff (nurses,
counselors, speech therapists) all accessed similar information in making classi-
fication decisions about a hypothetical case, regardless of the reason for referral.
Specifically, intelligence and achievement test data were chosen most often from
data available on intelligence, achievement, perceptual-motor, personality, lan-
guage tests, adaptive behavior scales, several behavioral observations, and be-
havioral checklists. Further, Ysseldyke et al. ( 1981) found in their study that the
reason for referral ( academic vs. behavioral problems) represented the most
critical variable in making classification decisions even when all available psy-
chometric data reflected average or normal performance.

In an attempted replication and extension of the Ysseldyke et al. ( 1981) study,
Huebner and Cummings ( 1985) found contrasting results. Specifically examin-
ing the decision making of school psychologists alone, Huebner and Cummings
( 1985) showed that school psychologists made highly accurate classification
decisions. Although the results of this study casts some doubt on the Ysseldyke

-2-