• Requires same skills as good teaching

1. Know the subject matter
2. Have clear learning outcomes
3. Know your students
4. Have good judgment
5. Be persistent
6. Have a touch of creativity
• Requires knowing a few simple technical skills

OH 2
Short-Answer and Completion Items

• Essentially the same:
• Direct question vs. incomplete sentence

Uses?

• Simple outcomes mostly (knowledge)
• Terminology
• Specific facts
• Principles
• Method or procedure
• Somewhat more complex outcomes
• Interpret diagrams, charts, graphs, etc.
• Very complex outcomes (math and science mostly)
• Answers can be given in numbers/symbols
• Answers can be provided on a scantron grid
• Where normally would use multiple choice, but options would give clues

OH 3
Short-Answer and Completion Items

Advantages?

• Easiest to construct
• Students must supply answer (partial knowledge not enough)

• Unsuitable for most complex outcomes
• Difficult to score (partly right answers, right but off-target answers, poor spelling)
• (Therefore) less objective and more time-consuming to score

OH 4
Short-Answer and Completion Items

Guidelines for writing?

1. Word question so that answer will be brief (word, phrase, number, symbol)
2. Word question so that only one answer is correct (e.g., at the right level of generality)
3. Don’t take wording directly from textbook (usually too general, vague, or provides simple association for student)
4. Use direct question rather than incomplete sentence (more natural, less ambiguous)
5. If answer in numerical units, say which ones and what level of precision required (except when that is part of the task!)
6. Use blanks of equal length and put to right side of questions
7. Don’t use too many blanks in one item (meaning gets lost)
8. Review the checklist for writing such items (p. 178)

• Single declarative statement
• Only one of several ways to respond: T/F, Y/N, F/O, T/F/O
• Variation on multiple choice
• Can require correction of false answers (with completion)

Uses?

• Measure knowledge
• Measure understanding (limited)
• Generalize to new situations
• Fact-opinion (F/O)
• Cause-effect
• Simple aspects of logic (e.g., if-then)

• Time-efficient (3 T/F per 2 multiple choice)
• Easy to construct (but more illusory than real!)
• Can measure a broad range of knowledge
• Can sample wide range of content (in some subjects)
• Objective scoring

Limitations?

• Not useful in many subjects
• Not useful beyond knowledge outcomes (except for F/O, cause-effect)
• Susceptible to guessing. Therefore:
• Many students get most items right
• Not clear what knowledge reflected
• Corrections for guessing a problem.
• Corrects only for random guessing, not clues
• Favors risk takers
• Reliability low per number of items
• Items not diagnostic when missed
• Student response sets will increase unreliability when T/F answers not balanced

OH 7
True-False Items

1. Understand that aim is to avoid ambiguity and clues—and that is hard!
2. Avoid broad general statements (for T/F). Most generalizations are false, and qualifiers give clues.
3. Avoid trivial statements (minutiae)
4. Avoid negative statements, especially double negatives
5. Underline any negative words (not, etc.)
6. Avoid long, complex sentences
7. Avoid two ideas in one statement unless a cause-effect item (i.e., leave out non-essential clauses)
8. Avoid non-attributed opinions, unless the ability to spot opinion is being tested
9. Avoid T/F items that unequal in length (longer are usually true)
10. Avoid highly disproportionate number of T or F answers (favors particular response set)
Review checklist (p. 185)

• Variation on multiple choice (a combined set of them)
• Items are homogeneous in content (same topic)
• Elements in the column for which a match is sought are premises
• Elements in the column from which selections are made are responses

• Measure knowledge (dates, persons, principles, etc.)
• Task is to associate two things on a logical basis
• Limited value overall
• Best to start with multiple choice and switch only when they turn out to be homogeneous and thus repetitive in form

• Compact and efficient way to measure knowledge
• Easy to construct (but more illusory than real! Hard to have plausible but wrong options)
• Objective scoring

Limitations?

• Susceptible to giving clues (only T/F worse)
• Restricted to simple factual learning
• Difficult to find enough homogeneous items that not trivial

OH 10
Matching Items

1. Use only homogeneous items
2. Remember that items must be more homogeneous at higher levels of development
3. Use unequal number of responses and premises (reduces guessing)
4. Say (and mean!) that responses can be used from zero to several times
5. Keep list of items to be matched brief (4-7 usually, and never more than 10)
• Aids homogeneity
• Allows to test to provide more balanced coverage
• Easier for students to scan full set
6. Put list of shorter responses on the right (easier for students to scan alternatives)
7. Arrange responses in logical order (alphabetic, date)
• Easier to scan
• Fewer clues
8. Directions should state basis of match
• Saves students figuring that out
• Avoids confusion, ambiguity
9. Directions should not be long and involved
10. Put all items on same page (no flipping, no missing)
11. Review the checklist for writing such items (p. 190)