There is an extremely large number of standardized tests administered
in this country. Most of them will fall into one of three categories:
Achievement tests, Scholastic Aptitude tests, and Specific Aptitude
tests. These three categories are summarized in the following table.
Kind of Test
|
Purpose
|
Reliability
|
Special Considerations
|
Achievement Tests
(ACT) |
To assess how much
students have learned from what they have specifically been taught.
|
- Reliability
coefficients are often .90 or higher; they are typically higher
for secondary students than for elementary students. Coefficients
may be somewhat lower for subtest scores.
- Content validity
must be determined for each situation.
|
- These tests are
usually more appropriate for measuring broad areas of achievement
than specific knowledge or skills.
- Useful when
comparing the performance of two groups of students.
- Allows us to follow
a student's progress over time and can alert us to "potential
trouble spots".
|
Scholastic Aptitude
Tests
(SAT) |
To assess students’
general capability to learn; to predict their general academic
success over the short run Often called intelligence tests.
|
- Reliability
coefficients are often .90 or higher; they are typically higher
for secondary students than for elementary students.
- Predictive validity
for academic success ranges from .40 to .70, depending on the
situation and student population.
|
- Test scores should
not be construed as an indication of learning potential over the
long run.
- Individually
administered tests (in which the tester works one-on-one with a
particular student) are preferable when students’ verbal skills
are limited or when significant exceptionality is suspected.
- Typically assess how
much a student has learned from their everyday lives and
experiences.
|
Specific Aptitude
Tests
(PSAT) |
To predict how well
students are likely to perform in a specific content domain.
|
- Reliability
coefficients are often .90 or higher.
- Predictive validity
for academic success often falls below .50.
|
- Test scores should
not be construed as an indication of learning potential over the
long run.
|