Kinds of Tests


Standardized tests have been developed by commercial companies and evaluated through the use of a carefully chosen group of test-takers that represents a wide-range of “typical” test takers that are expected to use the test.

Results are statistically “manipulated” to find “average” scores (the “norm”) for specific sections of the test. Test administration conditions are carefully controlled or standardized to ensure that each time the test is given, results will accurately demonstrate how the test taker has performed when compared to the “norm” group. These tests typically are both

  • reliable – they should be expected to show consistent results for each child, and 
  • valid – the tests should be accurately testing the content that they are supposed to test 

Use of a standardized test permits a valid comparison among scores from other tests that also are “standardized.” 
Norm-referenced scores compare a child to the “norm-group” used to develop the test. For assessing special-needs students, one must be sure that the test developer included children who shared the special-needs child’s characteristics in choosing a norm group.


Criterion-referenced tests are scored according to a predetermined ABSOLUTE standard based on decisions by the test maker.

  • Example: A criterion-referenced test score of 100 might mean “perfect score” (all items tested were answered correctly). 

The level to be considered “mastery” or “satisfactory” is determined by those who develop the test.

  • Example: 85% correct might be called a passing score. 


A benchmark test measures knowledge of the content of a certain text or curriculum at predetermined points (benchmarks) in the text or curriculum.  Students in the public school classrooms must pass specific benchmarks in order to move ahead.


A performance assessment (or performance-based assessment) is an evaluation of a child’s capability to carry out selected tasks based on recent learning. Such assessments are usually not a standardized paper and pencil type of measurement, but they are helpful in gathering an ongoing record of the child’s work on classroom learning tasks.
Many performance assessments use a “rubric” to guide grading of the work. A rubric systematically states the standards for the components of the task being done. For example, a penmanship rubric would have a standard for staying on the line, letter form and size, spacing between letters, and legibility. 


Portfolio assessments are particularly useful for the special needs child who may not be able to carry out more difficult tasks on his or her current grade level. The portfolio is a collection of school work products collected over a specific period of the school year. They are evaluated based on rubrics or on demonstrated progress.


Teachers conduct informal and formal tests as they teach. In many cases, the teacher is assessing student learning simply by asking an oral question. At other times, the teacher may ask the student to do something  (such as write a problem or answer on the whiteboard.) Teachers typically set standards for a passing score on a specific selection of content covered that will challenge the most able students, yet permit the least able to show their level of accomplishment as well.