maximum performance - procedures and conditions are set for the student to attain the best score he/she can.

typical performance - procedures and conditions are set so the student will perform under ordinary or typical conditions.

1. When it will be given.
2. Under what conditions it will be given.
3. The general content areas the assessment will cover.
4. The weighting of each content area.
5. The types of performance the assessment will require.
6. How the assessment will be scored/graded.
7. The importance of the assessment to particular decisions.

In relation to the above bits of information students need in order to perform their best:

• students need enough notice to prepare in advance,
• the notion students should always be prepared is foolish, and pop quizzes are not recommended,
• test anxiety and fear are reduced when students have at least 48 hours notice of an assessment,
• students need to know how many items, how much time, whether it is open- or closed-book, if there is a penalty for guessing, and what time of day it will be given,
• students need detail to study effectively,
• weight of different parts of the assessment should match the emphasis given it during instruction,
• students should have the opportunity to practice the performance that will be tested,
• knowing how an assessment will be scored helps students to prepare accordingly, and
• students need to know what will be done with the results.

Students also need the following minimum test-taking skills:

1. How to pay attention to directions and what the consequences will be if they fail to follow them.
2. Finding out how the test will be scored, how individual tasks will be weighted, and how many points will be deducted for wrongs answers, misspellings, and poor grammar.
3. Ability to write/mark answers clearly to avoid lowered scores.
4. How to study throughout the course and review appropriately.
5. How to use assessment time wisely.
6. How to use partial knowledge and guessing appropriately.
7. Ability to reflect, outline, and organize essay answers before writing them.
8. Remembering to double-check marks made on separate answer sheets.
9. How to review and change answers of there is a better response.

Certain clues in items will alert test-wise students to the correct answer (e.g., obvious associations between words in stem and words in an alternative, specific determiners eliminating answers, longer and more qualified answers being the key, grammatical clues in the stem, and overlapping alternatives).

test-wiseness - the ability to use test-taking strategies, clues from poorly constructed items, and past experience to improve one's score beyond what it would normally be.

Although you should construct or use top-quality assessments, you should also make sure all students have the skills below to put them on an even par with more test-wise students:

1. time-using strategy
2. error-avoidance strategy
3. guessing strategy
4. deductive reasoning strategy
5. intent consideration strategy
6. cue-using strategy

Although the popular belief is that changes answers leads to making more mistakes, if the student uses thoughtful consideration, research shows that 2 out of 3 changed answers will be correct.

• lower ability students benefit less and more difficult items have less of a payoff.

Test anxiety is common in those students who are motivated to do well.

Task-directed thoughts are held by students who perceive evaluation as a challenge, where their thoughts and actions are focused on completing the task at hand to reduce any associated tension.

Task-irrelevant thoughts are commonly held by students who perceive assessments as threats, where they are self-preoccupied and center on what would happen if they failed, their own helplessness, and a desire to escape from the tension-producing situation as soon as possible (cognitive interference).

Factors that may be under the instructor's control include:

• students perceiving the test to be difficult,
• at-risk students having higher levels of test-anxiety,
• giving item-by-item feedback after the assessment,
• arranging items from easy to difficult,
• frequent testing reduces improves performance for the test anxious,
• instructions to concentrate attention on the task reduces distractions,
• test-wiseness training for students with low test-taking ability can improve their performance.

The appearance and layout of an assessment is vital for validity of the interpretations.

You should type or write the assessment items neatly, and not give them orally.

If there are a lot of objective items, an answer sheet should be provided and items should be placed in a booklet.

Arranging items by format also reduces students having to change their "mindset."

I can be beneficial to arrange content areas according to how they were taught in class.

Within each content area, items should be arranged from easy to difficult.

Test directions should contain minimum information such as the number and format of items, the time allotted, where and how answers should be indicated, what the penalty is for guessing, and general strategies the students should follow.

Tests should be copied with high quality and should be kept in secure locations.

Before administering a test, every keyed response should be verified.

Correction for guessing formulas can be applied to scores on multiple-choice and true-false tests (see formulas and variations on p. 318.

If standardized tests are used, the test manual must be followed or else the interpretations of the results are invalid.

Item Analysis

item analysis - the process of collecting, summarizing, and using information from students' results to make decisions about each assessment task/item.

Uses of item analysis for classroom tests include:

1. Determining whether an item functions as you intended.
2. Feedback to students about their performance as a basis for class discussion.
3. Feedback to the teacher about student difficulties.
4. Areas for curriculum improvement.
5. Revising the assessment tasks.
6. Improving item-writing skills.

Steps for doing an item-analysis are listed in Figure 14.8 and will be reviewed in class, as will the item analysis computations.

item-difficulty index - the fraction of the total group answering an item correctly (overall item difficulty for a complete test should average .50, with necessary adjustments for guessing).

item-discrimination index - the difference between the fraction of the upper group answering the item correctly and the fraction of the lower group answering the item correctly.

Item analysis for constructed-response and performance exams is similar to that of objective exams (see the formulas on p. 323).

Item difficulty can be used to:

• identify concepts that need to be retaught
• provide clues for possible weaknesses and strengths in school curricula
• give feedback to students
• provide clues for possible item bias

Item discrimination is important when you wish to rank order students (i.e., show that some are outperforming others).

More weight should be given to the discrimination index than to the difficulty index because the discrimination index can change the rank ordering from the way you intended, but the difficulty index does not.

Generally speaking, you want discrimination indexes to be positive, because a negative index would indicate that the lower scoring students had a greater percentage answering the item correctly than the upper scoring students.

Improving Multiple-Choice item Quality

Every distractor should have at least one student from the lower-scoring group selecting it, and more students overall from the lower group should select the distractors than students from the upper group.

Note that not every lower-scoring student will lack knowledge of the correct answer for a particular item, and not very higher-scoring student will have the correct answer indicated.

If no one from the lower group selects a distractor, it should be revised or eliminated.

Higher-scoring students will sometimes select an answer that was not keyed but was plausible, which would call for revision of the alternative (ambiguous alternative).

When lower-scoring students are divided between two alternatives, this is more likely to mean the students are selecting a common misconception, which is what you would expect from less knowledgeable students.

You need to always consider the characteristics of the group of students taking the exam and not just rely totally on the statistics to make revisions.

If a large number of upper-scoring students select a wrong answer, it could be that it was keyed incorrectly (miskeyed item) .

If each alternative is selected by an equal number of upper-scoring students, that would very likely be an indication of blind guessing due to lack of knowledge or a poorly written item (always look at patterns of the upper group and not the lower group to determine guessing).

Guessing adds to the standard error of measurement and decreases the reliability and validity of the interpretations of an assessment.

Items that perform well can be added to an item bank or pool for future use.

Even if items have difficulty and discrimination indices that are less than ideal, you should still select items that will cover the important areas of content outlined by the test blueprint.

Finally, for tests that measure only one ability, the difficulty index for items should be between .16 and .84.

For tests that measure more than one ability, the difficulty index for each item should be between .40 and .60.

If tests are used for rank ordering (as most classroom tests are), then the discrimination index for each item should be above .00.

If tests are used for absolute (i.e., mastery) achievement, then item discrimination indices should be above .30.