3.5 Educational Assessment and Achievement Testing: Norm-Referenced, Criterion-Referenced, and Intelligence Tests

Educational Assessment and Achievement Testing: Norm-Referenced, Criterion-Referenced, and Intelligence Tests

Quick Answer: The EIPA Content Knowledge Standards devote a full sub-topic to testing and assessment. The facts most often tested: many states require annual academic achievement tests, and often the instructions can be interpreted, but not the actual test — though it may be appropriate for an interpreter to interpret a test if the goal of the test is to assess content knowledge and not literacy. Without annual assessment, it is difficult to know how much progress a student has made. The central validity problem is that most standardized tests were developed and standardized with hearing students, so some items may not be appropriate for deaf or hard of hearing students and may not reflect their underlying abilities. Criterion-referenced tests use target skills a student is expected to have mastered by a given age. Intelligence tests attempt to measure cognitive abilities and processing strategies, and intelligence tests that use language often underestimate the intelligence of a deaf or hard of hearing student. Classroom checklists are generally not standardized, so the person completing one must be knowledgeable for it to be effective.

---

1. Why Assessment Is an Interpreter Competency

Interpreters do not write, score, or interpret the results of assessments. So why does the EIPA test this?

Three reasons, all practical. First, the interpreter is frequently in the room when a test is administered and must know what is and is not permissible to interpret. Second, the interpreter sits on the IEP team, where achievement data drives goals and placement decisions, and a team member who cannot distinguish a criterion-referenced measure from a norm-referenced one cannot evaluate the claims being made. Third — and this is the one the standards press hardest — test results for deaf and hard of hearing students are systematically vulnerable to misinterpretation, and an interpreter who understands why is positioned to raise it.

---

2. Four Kinds of Measures

Achievement tests are used to determine a student's improvement in reading, writing, and other content subjects. The standards defend annual testing on a simple ground: without annual assessment, it is difficult to know how much progress a student has made. Criticizing the instruments is not the same as arguing for measuring nothing.

---

3. The Standardization Problem

This is the most conceptually important idea in the sub-topic, and it is worth stating precisely.

A norm-referenced test reports a student's performance relative to a norm group — the sample on which the test was standardized. The standards note that a major problem with most standardized tests is that they have been developed and standardized with hearing students. Two distinct consequences follow:

1. Comparison validity. A deaf student's percentile is being computed against a group whose access to incidental language, classroom audio, and English print exposure differed fundamentally from the student's own.
2. Item validity. Some items may not be appropriate for students who are deaf or hard of hearing and may not reflect their underlying abilities. A reading item built on a rhyme, a homophone pun, or a phonological awareness task measures something other than comprehension for a student who does not access English through sound.

The takeaway the standards intend is not that test scores are worthless, but that a low score is ambiguous evidence. It may reflect content knowledge, or English reading ability, or the test's construction — and the team must not collapse those into a single conclusion about the student's capacity.

The intelligence test case is the sharpest version of this. Intelligence tests attempt to measure cognitive abilities and processing strategies, and the standards state that intelligence tests that use language often underestimate the intelligence of a deaf or hard of hearing student. That is exactly why nonverbal cognitive measures are the standard of practice for deaf learners, and why the historical record of misclassification based on language-loaded IQ testing is a recurring theme in deaf education. It also aligns with the cognitive-development standard that students who are deaf or hard of hearing have the same capability for cognitive development as do students with normal hearing — the capability is equal; the measurement instruments often are not.

---

4. What an Interpreter May and May Not Interpret During a Test

The operative standard: many states require annual academic achievement tests. Often the instructions can be interpreted, but not the actual test. It may be appropriate for an interpreter to interpret a test if the goal of the test is to assess content knowledge and not literacy.

The distinguishing question is what the test is measuring:

Two governance rules sit on top of this analysis, and the test expects you to apply them rather than improvising:

• The decision is not the interpreter's to make alone. Testing accommodations are decided by the IEP team and documented in the IEP, and state testing programs publish their own allowable-accommodation lists. An interpreter who decides at the desk to interpret a reading passage has invalidated the assessment.
• Assessments for deaf and hard of hearing students must be conducted in the student's native language and desired mode of communication, and IDEA requires that students who are deaf or hard of hearing receive a comprehensive communication assessment as part of the annual IEP review. Interpreter access during evaluation is therefore a procedural requirement, not a favor.

[!CAUTION] Exam Trap: "The student has an interpreter, so the test is accessible." Interpreting a reading test does not accommodate the student; it changes what the test measures. Conversely, refusing to interpret anything — including instructions — during a content-knowledge test can deny access the IEP team has authorized. Both errors appear as distractors.

---

5. Reading the Results Without Overreading Them

Interpreters sit at IEP tables where numbers get discussed loosely. A few habits keep the discussion honest:

• Ask what the number is referenced to. A criterion-referenced result ("mastered 7 of 10 target skills") and a norm-referenced result ("18th percentile") answer different questions.
• Ask what language the test was in. A content score produced by a heavily text-dependent test for a student reading well below grade level is at least partly a reading score.
• Treat checklists as opinions, not data. Because classroom checklists are generally not standardized, their value rests entirely on the knowledge of the person completing them.
• Contribute only what you can observe. The interpreter's IEP-team contribution is observations about how well the student understands the interpreted classroom and about the limitations of the interpreting process — not an evaluation of the student's academic or behavioral performance except as it relates to interpreting.

---

6. Realistic K-12 Scenarios

Scenario A: The Reading Subtest

A proctor hands the interpreter a state reading comprehension booklet and says, "Just sign the passages for him so he understands them."

• Analysis: The construct being measured is the student's ability to read English. Interpreting the passages substitutes signed comprehension for reading comprehension and invalidates the result.
• Correct action: Interpret the instructions and procedures, decline to interpret the passages or items, and refer the question to the testing coordinator and the IEP team. If the student's IEP authorizes a specific accommodation, that authorization — not a proctor's verbal instruction — governs.

Scenario B: The Science Unit Test

A 6th-grade science teacher gives a unit test on ecosystems. The deaf student knows the content but reads two years below grade level.

• Analysis: The goal of this test is content knowledge, not literacy. The standards state it may be appropriate for an interpreter to interpret a test in exactly this circumstance.
• Correct action: Confirm with the teacher and the IEP team in advance, so that the support is planned and documented rather than improvised mid-test — and so the same practice is applied consistently.

Scenario C: The IQ Score in the Eligibility Meeting

An eligibility report shows a deaf 3rd grader with a Verbal Comprehension index in the low range, and a team member concludes the student "has limited cognitive ability."

• Analysis: Intelligence tests that use language often underestimate the intelligence of a deaf or hard of hearing student, and the standards elsewhere affirm that deaf and hard of hearing students have the same capability for cognitive development as hearing students. A language-loaded index is measuring English language access as much as reasoning.
• Interpreter's role: Stay inside the lane the standards define. The interpreter does not reinterpret psychometric data, but can report factual observations about the student's comprehension of the interpreted classroom, and the team can request nonverbal cognitive measures administered by a professional with training specific to students who are deaf or hard of hearing.

---

7. Exam Traps

• Trap 1: Assuming nothing on a test may be interpreted. Instructions generally may be, and a content-knowledge test may be.
• Trap 2: Assuming everything may be interpreted. A reading test is the paradigm case where interpreting the items destroys the measurement.
• Trap 3: Reading a low standardized score as a statement about ability. Most standardized tests were normed on hearing students, and some items may not reflect a deaf student's underlying abilities.
• Trap 4: Treating a classroom checklist as standardized data. It is not; its usefulness depends entirely on the knowledge of whoever completed it.
• Trap 5: Concluding that annual testing should simply be abandoned. The standards defend annual assessment on the ground that without it, progress cannot be known.