A Terse Self-Test about Testing From Popham (2005), Page 3.

A Terse Self-Test about Testing

From Popham (2005), Page 3

1.The chief reason that teachers should give classroom tests is to determine students’ grades:

Strongly AgreeAgreeUncertainDisagreeStrongly disagree

While the use of tests to determine grades is an important use of tests it is not the only, or even the most important use of tests.

Other important uses of tests are:*Diagnose students’ strengths and weaknesses,*Monitoring students’ progress,*Determining instructional effectiveness,*Influence perceptions of educ’nl effectiveness,*Evaluating teachers,*Clarification of instructional intentions.

2. Teachers should typically plan instruction that focuses on the skills or knowledge represented by a test.

Strongly AgreeAgreeUncertainDisagreeStrongly disagree

First consider the reason for giving a test in the first place.

Why give a test that is NOT aimed at assessing important instructional/learning targets?

If we know, beforehand, what we want students to know and be able to do AND we are able to determine what we would accept as evidence that students have attained this knowledge than why not aim instruction at that knowledge?

A test is nothing more than a vehicle for collecting the evidence we seek.

3. In their classroom tests, teachers should only use items that can be scored objectively:

Strongly AgreeAgreeUncertainDisagreeStrongly disagree

Objective tests (e.g., multiple-choice tests, matching tests, true-false tests, short-answer, and fill-in-the-blanks tests) can be used, effectively, in a wide variety of situations.

But there are many situations, like assessing students’ ability to argue persuasively, or their facility in speaking a foreign language, that require teacher judgement.

4. There are other legitimate indicators of a teacher’s instructional effectiveness besides students’ test scores:

Strongly AgreeAgreeUncertainDisagreeStrongly disagree

The current trend, especially in light of NCLB, is to evaluate teachers’ instructional effectiveness in terms of students’ test scores.

However, several reasons can be given for why this may be inappropriate (these will be discussed later in this course).

Teachers can influence students, positively, in many ways that are not amenable to assessment via test scores.

5. A teacher has no business measuring students’ confidence in their ability to do school work:

Strongly AgreeAgreeUncertainDisagreeStrongly disagree

There are those who would argue that assessing students’ confidence in their ability is irrelevant or, at least, unimportant.

On the other hand, there is ample research that shows that students’ confidence in their own ability (often called self-efficacy) is strongly related to their performance.

6. Nationally standardize achievement tests should never be used to supply evidence about how well teachers are instructing children:

Strongly AgreeAgreeUncertainDisagreeStrongly disagree

Nationally standardized tests are designed, mainly, to assess student achievement relative to national norming groups. They are not particularly sensitive to individual differences in instruction or to differences in groups being instructed.

Mismatches can, and often do, occur between what is tested and what is taught.

Important teacher-stressed content is often excluded.

Attribution for good (or poor) performance on these tests is often questionable.

7. Teachers really do not need to determine the reliability of their own classroom tests:

Strongly AgreeAgreeUncertainDisagreeStrongly disagree

It certainly is important for teachers to know what reliability is, and to have a basic understanding of how it is computed and interpreted--especially since they are often called upon to explain standard tests results.

However, other that the fact that it can impact validity, reliability plays a relatively small role in classroom tests (most have low reliability anyway).

Teachers generally collect, and base educationally-relevant judgments on, a large assortment of assessment data.

8. It is impossible to judge the quality of students’ written compositions with any meaningful accuracy:

Strongly AgreeAgreeUncertainDisagreeStrongly disagree

Judging written work obviously involves teacher judgment, which by its very nature is subjective.

Often, attributes and characteristics of the written work that are irrelevant to the response requested influence the judgment.

However, with the careful use of well-designed rubrics, much of the capriciousness involved judging written work can be eliminated—or at least curtailed.

Judges can be trained, easily, to evaluate written work.

9. The enormous pressure to boost students’ scores on important tests permits teachers to employ almost any sort of score-improvement preparation alternatives:

Strongly AgreeAgreeUncertainDisagreeStrongly disagree

There are test-preparation procedures and techniques that can be considered educationally and ethically defensible and there are techniques whose educational and ethical defensibility is questionable.

One question to ask: Is performance on the test a consequence of learning or is it a consequence of the test-prep procedure?

10. Significant classroom tests should almost always be constructed prior to a teachers instructional planning:

Strongly AgreeAgreeUncertainDisagreeStrongly disagree

If important classroom tests are designed to assess significant learning outcomes (learning targets), then designing such tests requires that the teacher has clear understanding of what is to be taught and, more importantly, what the student should know and be able to do. By deciding beforehand what evidence of accomplishment is acceptable the teacher is in a good position to design the instruction.