"The Test Drive" yields many different meanings. The most obvious one refers to the testing of a new vehicle. We 'test drive' a car in order to see if it is "right" for us, if the 'feel' of the car is right. But test drives may also refer to test crashes, where we are trying to find the limits of the car. How much force can it take before it breaks? Its only by smashing the car that we are able to assess a certain quality.
We can also understand the 'test drive' as a drive to test. This seems to be a drive common to human beings--we want to try things out and we want 'tests' to confirm that it will act in such and such a way again. We claim that if something is "testable" it can be proven. If it 'passes the test' then a degree of truth can be found (so we usually think).
Ronell is interested in playing with (and interrogating) of the valences of "to test," including a "test" in the educational/pedagogical sense. Although she does not go into depth with this, it may be a way into looking at her argument in terms of the university.
While a scientific test has to be taken many times over in order to confirm that it has provided "proof" for knowledge, tests in schools only have to be taken once. As Ronell puts it, "even if a test, to fulfill its bald constative claims, assumes the function of providing providing definitive results or minimally of confirming that cognition occurs, testing, for its part and imprating, is always temporally determined" (186). This is basically common sense for students. Students take the test and once they have "proven" that they have grasped the information quickly throw it in the recycle bin of their minds. Thus, a 'test' in this sense is not really a "prototype" that will eventually form into something executed. Rather, it remains a prototype--a playing at knowledge rather than an enacting of it: "The normatively secured test does not generate knowledge but confirms what already exists as 'knowable'" (187).
Ronell asks whether it is possible for a test to generate knowledge. She argues that only in the test's failure does it generate knowledge and, furthermore, that this is the "unpretended aime of a test." She wrties, "Generous failure, productive disclosure, concerns a type of testing that probes more than the workability or conformity of its object to an already regulated norm" (188). By definition, 'standardized' tests cannot produce knowledge. It is knowledge-telling.
I'm not sure we need Ronell for the basic arguments I am making here against standardized testing, but its the scene that came to mind.