Page 130 - Making Instruction Work
P. 130

chap 11  3/11/97 4:59 PM  Page 116




             116                making instruction work


                   You:     Oh? What kind of tests?
                   Surg:    Mostly multiple-choice. But there were
                            some essay items, too.

                   You:     Good-bye!

                In practice it isn’t  always  possible for your test items to
             duplicate the conditions called for by the objective. In such
             cases one approximates those conditions as closely as possible.
                But it is always possible for a test item to demand the same
             performance as that described in the objective. For example, if
             an objective asks students to be able to repair equipment under
             water or to splice cables on top of a pole, it may not be possible
             to provide the water or the pole. In those instances you would
             provide the closest approximation to those conditions that you
             can. But you would always ask them to repair, and you would
             always ask them to splice. The rule is this: If you must,approxi-
             mate the conditions, but never approximate the performance.


             How to Do It

                Prepare a criterion test for  each  objective whose accom-
             plishment you want to measure. Many of those tests will con-
             sist of only one item (question), and the rest will need only
             three or four. How do you know how many items to include?
             The rule is this: The test will contain as many items as are
             needed to sample the range of conditions called for in the
             objective. Here are the steps for preparing a criterion test (skill
             check).
                1. Read the objective and identify the performance (what it
                   wants someone to be able to do).

                2. Draft a criterion item that asks students to exhibit that
                   performance.
   125   126   127   128   129   130   131   132   133   134   135