Summary

In this chapter, you have seen that two competing paradigms exist for classifying and categorizing language tests. The discussion of these paradigms necessitated defining a number of different concepts including norm-referenced tests, Criterion-referenced tests, domain-referenced tests, and objectives-referenced tests. For the purposes of this book, it was stated that Criterion-referenced tests would be used as a cover term which would include the variant sampling procedures represented by domain-referenced and objectives-referenced tests.

The chapter then went on to explore the differences between the two main categories of tests: CRT and NRT. It was argued that CRTs were developed in response to a number of problems with using NRTs for some types of educational decisions and that the advent of CRTs answered many of those problems.

The chapter ended with a discussion of the place of CRTs in language testing. In the quest to discover what language tests are measuring, four questions were addressed:

1. What makes language testing special?
2. What is language proficiency?
3. What is communicative language ability?
4. And, what problems do CRT developers face?

This chapter addressed a number of overarching issues in language testing. Though the general orientation was to explain the characteristics and benefits of Criterion-referenced tests, theories of language proficiency and communicative competence were necessarily discussed as they relate to Criterion-referenced testing. However, no attempt was made to claim that any one of the theories was better than any other. We have merely attempted to show how the test developers view of any particular issue will affect decisions on what to test and how to interpret test results.