News story

Online Language Testing - Conciliating Efficiency with Validity

Online Educa Berlin 28-30 Nov 2007 - ExhibitorLearning NewsMondiale Testsysteme

More than any other form of online testing, language testing is confronted by two diametrically opposed demands. The requirement for validity on the one hand and the need for efficiency on the other.

Validity in online testing covers a number of different topics beginning with the content and scope of the test items, extending to repeatability and comparability of the results and finally including issues of security like preventing the use of unauthorized resources (cheating) and authentication of the test candidate; insuring that the candidate is actually the person he purports to be.

The assessment of online tests achieves it's efficiency through objective or algorithmic evaluation and for many subjects this approach works well. Language listening and reading comprehension skills lend themselves well to objective evaluation but creative skills such as speaking or writing can only be adaquately assessed subjectly, requiring human intervention and potentially impeding efficiency.

The objective of successful online language testing is to attain an acceptable degree of validity at the same time maintaining a level of efficiency commensurate with the expectations of online testing. There are many different types of tests for different objectives but MONDIALE-Testsysteme only offers the test types most widely known online: placement and proficiency tests. Placement tests are those taken prior to participating in a language training course to assign the new student to the appropriate group. Proficiency tests either test the achievements of a participant in a training program or assess the language skills of an applicant for admission to an institution, employment or transfer. Higher standards must be applied to proficiency tests. An important aspect of validity in the context of proficiency tests is that the results must be easy to interpret and comparable to assessments from other sources. In short they must be mapped onto an accepted standardized scale such as the Common European Framework for Languages.

Another variable in the validity equation is the granularity or the preciseness of the assessment scale. Although the CEF is a helpful tool for interpreting test results, in many cases it does not discriminate sufficiently well to compare test candidates. When a large number of individuals with similar skill levels must be differentiated, a more exact scale is necessary than for a small number with a broader range of skills. For instance several thousand international students applying for a limited number of openings at a university need to be rated on a much finer scale than a hundred employees being placed into language training classes. An efficient testing strategy must take this into account.

To reach the appropriate balance between efficiency and validity is no easy task. Above all there is no single correct strategy for all situations. MONDIALE-Testsysteme presents proposals for online placement tests and proficiency tests that are based on requirement-oriented validity standards.