Selected response item tests, as well as being popular methods of measurement, are thought to be reliable and objective. They also meet the two cardinal requirements regarding efficiency and practicality. None the less,traditional selected response item tests lack a sound basis in terms of item writing theory. This chapter details the application of schema theory to the construction of selected response item tests in order to provide such a basis.The findings demonstrate that schema-based selected response items constructed on authentic texts are reliable and correlate significantly with the granunar, vocabulary and reading sections of TOEFL. Furthermore, the results indicate that while native and non-native English speakers perform differently on schema-based items, they both regard the tests as a good measure of English proficiency. Thus, in addition to having criterion and face validity, schemabased selected response item tests are reliable, enjoy theoretical advantage over their traditional counterparts, help item writers dispense with constructing reading passages for the sake of testing rather than reading, provide the item writers with a rich source of item options, and take less time to be answered than tests deemed equivalents, such as the reading section of TOEFL.

Keywords