Reviewer
Reviewer
Reviewer
COMPLETION ITEMS
★ Items should require a single-word answer, or a brief and definite statement
★ Be sure the question poses a problem to the examinee
★ Be sure the answer is that the student is required to produce is factually correct
★ Be sure the language used in the questions is precise and accurate in relation to the subject
★ Omit only key words
★ Word the statement such that the blank is near the end of the sentence rather than near the
beginning
★ If the problem requires a numerical answer, indicate the units which it is to be expressed
WRITING ESSAY TEST ITEMS
★ An essay item is one for which the student applies, rather than selects, the correct answer
★ The student must compose a response to a question for which no single response can be cited as
correct to the exclusion of all answers
★ Essay items may be well constructed or poorly constructed
○ The well-constructed essay item aims to test complex cognitive skills by requiring the
student to organize, integrate, and synthesize knowledge
○ The poorly constructed essay item may require the student to do no more than recall
information as it was presented in the lecture or textbook
★ The potential of the essay item as an evaluation device depends not only upon writing appropriate
questions that elicit complex cognitive skills, but also upon being able to structure the situation
FORMULA
★ p = Number of students selecting correct answer
Total number of students attempting the item
★ D = (Numbers who got item correct in upper group) - (Number who got item correct in lower group)
Number of students in either group (if group sizes are unequal, choose the higher number)
★ When p levels are less than about .25, the item is considered relatively difficult
★ When p levels are above .75, the item is considered relatively easy
★ Test construction experts try to build tests that have most items between p levels of .20 and .80,
with an average p level of about .50
QUANTITATIVE ITEM ANALYSIS
★ Quantitative item analysis also enables us to make other decisions. We can use quantitative item
analysis to decide whether an item is miskeyed, whether responses to the item are characterized
by guessing, or whether the item is ambiguous
★ Miskeying - most students who did well on the test will likely select an option that is a distractor,
rather than the option that is keyed
★ Guessing - students in the upper half of the the class respond in more or less random fashion
★ Ambiguous - among the upper group, one of the distractors is chosen with about the same
frequency as the correct answer
★ Guessing Formula
Score = Total Right - Total Wrong .
No. of answer choices - 1
DEBRIEFING
Before handing back answer sheets or grades:
1. Discuss problem items
2. Listen to student reactions
3. Avoid on-the-spot decisions
4. Be equitable with changes
After handing back answer sheets or grades:
1. Ask students to double-check
2. Ask students to identify problems
In trying to remain non defensively and emotionally detached from your test, it is useful to keep a few
points in mind:
1. Your test will include at least some items that can be improved
2. Students are criticizing your skill as a test constructor, not you as a person
3. The quality of an item is not necessarily related to the loudness with which complaints are made.
At times, items that students loudly protest are indeed poor items