Machine summary:
"Unlike many traditional types of tests, such as multiple-choice questions, in which scoring is highly consistent and very easily done, rating in performance-based assessment becomes a major concern because raters deal with real-world performances, not the simple tests of factual knowledge (Shohamy, 1995) Hughes (2003) also provides an example of a composition writing test to clarify the tension between reliability and validity.
Unlike many traditional types of tests, such as multiple-choice questions, in which scoring is highly consistent and very easily done, rating in performance-based assessment becomes a major concern because raters deal with real-world performances, not the simple tests of factual knowledge (Shohamy, 1995).
As discussed above, in the section about the history of performance-based assessment, it should now be clear that this approach to language testing, or testing in general (see Robert, Eva, and Dunbar, 1991, for a variety of subject matters in which performance-based assessment is used), was basically an attempt to satisfy the needs of the governments or universities that were seeking to find ways to have a valid measure of people’s real abilities, rather than their mere knowledge.
The increasing number of the foreign students entering British and American universities constituted the practical reason for embracing performance-based assessment, and the flourishing of Communicative Language Teaching is considered as the theoretical basis of this new approach to testing."