Page 330 - DEDU504_EDUCATIONAL_MEASUREMENT_AND_EVALUATION_ENGLISH

Educational Measurement and Evaluation

Notes Several types of questions are asked in them. These questions are generally classified into two
types—Recall type questions and Recognition type questions. Recall type questions have two types—
Simple recall type and Sentence completion questions. Recognition type questions have serveral
forms in which are included Classified questions, True/False questions, Multiple choice type
questions, Matching type questions, Multiple choice and completion questions, and Analogy
questions. We shall clarify their form and difference here by examples :

28.2 Objective Test Formats

A variety of different types of objective test formats can be classified into two categories : a selected
response format, in which examinees select the response from a given number of alternatives,
including true/false, multiple choice, matching test items; and a constructed response format, in
which examinees are required to produce an entire response, including short answer test items.
This distinction is sometimes captured in terms of recognition and recall. These two general categories
are further divided into basic types of objective tests, illustrated in the following examples.
The true/false test is the simplest form of selected response formats. True/false tests are those that
ask examinee to select one of the two choices given as possible responses to a test question. The
choice is between true and false, yes and no, right and wrong, and so on. A major advantage of the
true/false test is its efficiency as it yields many independent responses per unit of testing time.
Therefore, teachers car cover course material comprehensively in a single test. However, one apparent
limitation of the true/false test is its susceptibility to guessing. It should be noted, however, that test
givers can attenuate the effects of guessing by increasing the number of items in a test. In addition,
guessing might reflect partial knowledge, which would provide a valid indication of achievement.

28.3 Construction of Objective Test Items

Basically, scoring objective test items is easy : It only requires one to follow the scoring rules. However,
constructing good objective test items requires much more skill and effort. The first step is to develop
a set of test specifications that can serve to guide the selection of test items. A table of specifications
(or test blueprint) is a useful tool for this purpose. This tool is usually a two-way grid that describes
content areas to be covered by the test as the row headings and skills and abilities to be developed
(i.e., instructional objectives) as the column headings.
After specifying the content and ability covered by the test using the table of specifications, the
appropriate test item format is selected for each item. At this point, not only objective test items but
also other types of test items—essay test or performance assessment—should be considered,
depending on the learning outcomes to be measured.
The next step is to create specific test items. Typically, it is particularly important for objective test
items to be written in clear and unambiguous language to allow examinees to demonstrate their
attainment of the learning objectives. If complex wording is used, the item simply reflects reading
comprehension ability. It is also important for each objective test item to focus on an important
aspect of the content area rather than trivial details. Asking trivial details not only makes the test
items unnecessarily difficult, it also obscures what the test constructor really wants to measure.
Similarly, relatively novel material should be used when creating items that measure understanding
or the ability to apply principles. Items created by copying sentences verbatim from a textbook only
reflect rote memory, rather than higher order cognitive skills.

Test constructors must be very careful that examinees with little or no content knowledge
cannot arrive at the correct answer by utilizing the characteristics of the test format that
are independent of specific content knowledge.

324 LOVELY PROFESSIONAL UNIVERSITY