Authenticity is regarded as an important feature of language tests, but commonly the notion is related only to the use of authentic material. In the course of this class, however, we found out that the concept of authenticity is actually far more comprehensive. It does indeed include characteristics of the input of test tasks, but it is also concerned with the interaction between input and expected response, the setting and the format of a test. Hence, we decided to work mainly on the topic of authenticity in this final report of our course on language assessment.
According to Charles Alderson (2000: 138) the goal of all reading assessment “is typically to know how well readers read in the real world”. Authenticity is thus an important aspect of testing since it describes the relationship between the test and the real world. The notion of authenticity was developed first within applied linguistics in the 1960s. During the late 1970s and 1980s conceptions of authenticity in language teaching and testing concentrated on the use of genuine texts that had not been simplified and tasks that simulated real-world tasks (Lewkowicz 2000: 45-46). The debate mainly focused on the nature of authentic input until 1991, when Bachman suggested that two types of authenticity had to be distinguished: situational and interactional authenticity (Lewkowicz 2000: 48). While situational authenticity was concerned with the relationship between the test task and the criterion behaviour, interactional authenticity indicated the interaction between test task and test taker.
This notion was developed further by Bachman and Palmer in 1996, who differentiated between authenticity and interactiveness as two separate concepts (Bachman & Palmer 1996: 23-29). They define authenticity “as the degree of correspondence of a given language test task to the features of a TLU task” (Bachmann and Palmer 1996: 23). In this quotation the abbreviation TLU stands for ‘target language use’; Bachman and Palmer define the TLU domain as the “situation or context in which the test taker will be using the language outside of the test itself” (Bachmann and Palmer 1996: 18).
Other definitions of authenticity are rather similar. The Dictionary of language testing, for instance, states that “a language test is said to be authentic when it mirrors as exactly as possible the content and skills under test” (Davies 1999: 13). McNamara (2000: 131) defines authenticity as “the degree to which test materials and test conditions succeed in replicating those in the target situation”. What all these definitions have in common is the fact that they stress the ‘degree’ of authenticity, which indicates first that there can be no such thing as a completely authentic language test and second that the degree of authenticity is to be located on a continuum.
For the evaluation of the level of authenticity of our class project, we propose to work with the definition of Bachman and Palmer (1996) quoted above and with their framework for the evaluation of usefulness (Bachman & Palmer 1996: 137-156). However, we will concentrate on the issue of authenticity only since any further considerations of test usefulness would go beyond the scope of this report.
To determine the degree of authenticity, minimum acceptable levels might be stated in two ways according Bachman and Palmer (1996: 136): “in terms of task characteristics [or] in terms of expected perceptions on the part of test takers and test users”.
As we do not have any empirical information on the latter, we will concentrate on the first aspect and follow Alderson’s (2000) adaptation of the Bachman and Palmer framework for assessing reading. Although Alderson rarely makes use of the term ‘authenticity’ in his book Assessing reading, many of his considerations actually concern the relationship between tests and the TLU domain. The following analysis of the level of authenticity of our class project reading test is therefore based on an enumeration of relevant task characteristics by Bachman and Palmer (1996: 49-50) and their interpretation for reading by Alderson (2000: 140-166).
The first category is called ‘characteristics of the setting’ and deals with influences of the physical and temporal test settings on authenticity. Our reading test was administered in a classroom, which is semi-authentic in our view. Students often read texts in class and reading in and for school is certainly part of their real-life reading behaviour. However, they obviously also read in numerous other places and situations and for many different purposes. In addition, in the testing situation the pupils could not choose a time for reading, they had to do it under certain constraints. Again, this is semi-authentic because it is not always possible to choose a time for reading in the real world either, especially in school or work-related contexts. A third aspect of the characteristics of the setting deals with the participants in the reading process. Here we found a marked difference between our reading test and real-life reading. While all eighteen texts of the test were anonymous without any sources given, outside the testing situation the author or at least the source of a text is often known (Alderson 2000: 144). Thus, in real life the knowledge or indeed lack of knowledge about the writer and their opinions might influence the reading process, which is clearly not the case for our test. In summary, the characteristics of the setting only correspond partly to TLU reading behaviour of students for some aspects and they do not replicate others at all.
- Quote paper
- Katharina Eder (Author), 2010, Authenticity in Language Assessment, Munich, GRIN Verlag, https://www.grin.com/document/172655