Information Technology Reference
In-Depth Information
of DiscoTests. First we describe the tests and how they are scored. Then we describe
how we build them, and how they can be used.
Anatomy of a DiscoTest.
Because teasers are designed to be tests in the sense
described above, they must (1) provide students with an opportunity to engage in
meaningful action on their knowledge and (2) offer useful feedback. These require-
ments forced us to rule out multiple choice items like the one shown in Fig. 10.3.
This item asks the student to select one of five possible responses. There is one right
answer. The other answers are intended to represent common misconceptions held
by students. A number of assumptions accompany items of this kind. For example,
it is assumed that students who get an item right either (1) know the answer or (2)
have made a good guess. Also, it is generally assumed that students who get an item
right without guessing know more than students who get an item wrong. If these
assumptions held, items of this kind might provide information that could inform
accurate feedback—but the assumptions do not hold.
Students who select the right answer (b) do so for several different reasons,
many of which reflect partial knowledge or misunderstanding. Here are some of
the explanations students have given for choosing the correct answer to the item in
the example:
1. The right pan will not move because the amount of matter in a closed con-
tainer remains the same no matter what chemical or physical changes take place
(textbook response, could be memorized).
2. The right pan will not move because a gas was formed but nothing was destroyed
(answer showing partial understanding).
3. The right pan will not move because not even bubbles can get out of a jar that is
closed tight (answer showing that the student believed the jar was closed really
tight).
4. The right pan will not move because the gas does not have any weight (answer
showing partial understanding).
5. The right pan will not move because in a chemical reaction, atoms rearrange to
make new substances, but none of them are destroyed (answer showing greatest
level of understanding).
This phenomenon, which has been described by numerous researchers (Sadler,
2000), strongly suggests that multiple choice items are unlikely to provide the
kind of information required to inform educative feedback. Sadler has shown how
multiple-choice tests can be used more effectively, but multiple choice items, no
matter how well they are constructed, still limit the learning functions of an assess-
ment. Consequently, DiscoTests are composed of items that are open-ended and
require short essay responses consisting of judgments and justifications that not only
show (1) what students know, but also (2) how they understand what they know, and
(3) how they can use their knowledge to deal with similar tasks and situations. The
item in Fig. 10.3, stripped of its multiple-choice options, can function as a DiscoTest
item, as shown in Fig. 10.4.
