Well-constructed W-SR test items help the user/designer reliably discriminate between masters and non-masters of targeted learning outcomes in the domain. The chapter detailed the guidelines for constructing four major types of W-SR items: Matching, Completion, True and False, and Multiple Choice. These four basic item types can be modified in various ways, but item writers should make every attempt to ensure that the items:

• measure important learning outcomes;
  
• elicit the desired performances in a straightforward manner;
  
• pose a single question or problem with one clearly correct or best answer;
  
• do not provide clues to testwise examinees; and
  
• follow specific guidelines applicable to the chosen W-SR format.
  

With appropriate attention to item-writing mechanics, all four W-SR formats could be employed by creative item-writers to measure higher levels of learning. However, matching and completion formats appear to be better suited for assessing factual knowledge (knowledge and comprehension in Bloom’s taxonomy) levels of cognition, particularly in younger or special populations. True/false and multiple choice formats are excellent for the measurement of higher level skills, including application, analysis, and evaluation processes.

Page 5