In terms of authenticity and closeness to real-world language use, tests which involve more than one skill are increasingly thought to be preferable. Such tests are, however, more demanding on the setter, and despite their high validity (see Module 13, section 13.1.4.1), they can lead to confused measurement, especially if performance on one element (eg a prior reading or listening task) interferes with performance on a subsequent speaking or writing task. The creation of appropriate criteria can also be more challenging.

One such integrated task for oral testing involves candidates in renarrating the content of something they have heard on tape. The mark scheme for this can be quite specific, listing the points the candidate has to reproduce, but it needs to be supplemented by criteria for the quality and range of language employed. An alternative is to give candidates a text to read for a few minutes, to take the text back and to ask them to renarrate or summarize the text in their own words. The integration of reading and listening skills involved here can be furthered if candidates are allowed to retain the text and the oral examiner then asks specific questions on aspects of the text - eg 'What does the phrase 'x' in the last sentence mean?, or: 'Can you explain in your own words what is meant by the second sentence?', etc.