Kane, M. T. (2013). Validating the Interpretations and Uses of Test Scores. Journal of Educational Measurement, 50(1), 1–73
I'm working through this paper and noting it up in Mendeley. Here I'm going to note some (perhaps tangential) observations. Linking the test of those wanting visas or citizenship to the CEFR effectively lets the test designer off the hook. If you say to them, "How does your test give the inference that the test taker is fit for leave to remain?" They are liable to reply, "We are saying no such thing, we're saying that the test taker has B1 level speaking in accordance with the CEFR." And they may be able to point to research which validates the test in those terms. And so we need to turn to the test's principal stakeholder, the UK government, and ask, how CEFR level B1 can be interpreted (in effect) as a score which measures fitness for leave to remain? Where is the validity argument? It's unlike, for example, an IELTS score of 6.5, which it is generally assumed fits an L2 undergraduate to study hard science in an English medium University...