1. Those in which the item writer is guided by statements of the objectives of
instruction.
2. Those whose items must meet specifications of the domain of content to be
covered and the forms of items to be used.
3. Those in which items are produced by linguistic transformations of segments
of prose instruction.
4. Those in which mapping sentences derived from facet theory are used to
define a content domain.
5. Those whose items are designed to test understandings of concepts.
6. Those in which items are stored in or actually produced by computers.
The limitations of these methods is acknowledged clearly in the review. Each
method appears to have a particular application. They cannot be applied to any
content level and at any cogn itive level. They require ingenuity and the exercise
of judgment. At present they are in the infancy of their development. Cronbach
believes that they will mature into useful tools for the test constructor. Others,
including this writer, are more skeptical. Roid and Haladyna endorsed Berk's
(1978) observation that the rigor and precision of item-writing specifications are
inversely related to their practicability.
In a sense, the item development procedures outlined in earl ier sections of this
chapter constitute a technology for item writing . The form and derivation of the
items is specified quite precisely. The content of the items depends on the item
writers' knowledge and ski ll s. Propositions that are important and defensible
must be selected. They must be expressed clearly, accurately, and concisely.
Incorrect answer options that have commonsense plausibility must be provided.
The judgment involved in these choices is crucial, and no algorithm or computer
program is likely to provide it.
1. Those in which the item writer is guided by statements of the objectives of
instruction.
2. Those whose items must meet specifications of the domain of content to be
covered and the forms of items to be used.
3. Those in which items are produced by linguistic transformations of segments
of prose instruction.
4. Those in which mapping sentences derived from facet theory are used to
define a content domain.
5. Those whose items are designed to test understandings of concepts.
6. Those in which items are stored in or actually produced by computers.
The limitations of these methods is acknowledged clearly in the review. Each
method appears to have a particular application. They cannot be applied to any
content level and at any cogn itive level. They require ingenuity and the exercise
of judgment. At present they are in the infancy of their development. Cronbach
believes that they will mature into useful tools for the test constructor. Others,
including this writer, are more skeptical. Roid and Haladyna endorsed Berk's
(1978) observation that the rigor and precision of item-writing specifications are
inversely related to their practicability.
In a sense, the item development procedures outlined in earl ier sections of this
chapter constitute a technology for item writing . The form and derivation of the
items is specified quite precisely. The content of the items depends on the item
writers' knowledge and ski ll s. Propositions that are important and defensible
must be selected. They must be expressed clearly, accurately, and concisely.
Incorrect answer options that have commonsense plausibility must be provided.
The judgment involved in these choices is crucial, and no algorithm or computer
program is likely to provide it.
การแปล กรุณารอสักครู่..
