As Brown (1995) points out,
language tests are normally used to help make decisions, and there are a number
of types of decisions that they can be used for. We generally refer to tests,
in fact, by the type of decision they are used to make. I think it is useful to
divide these test and decision types into two broad categories: those that are
closely related to a teaching or learning curriculum, and those that are not. I
use this distinction because curriculum-related tests all have a specific
domain—the curriculum—to which we can refer when planning and writing these
tests. In contrast, when a test is not based on a particular curriculum, we
have the burden or freedom (depending on one’s point of view) of deciding what
specifically
These types of tests are
summarized in Table 1.1. Brief consideration, of course, will show that many
tests are used for more than one purpose; I will refer to several common types
of overlap in the following discussion. This is not necessarily
problematic. it should be based
on.
Furthermore, as will become
evident shortly, the dividing line between one type of test and another is not
always as clear and sharp as we might pretend. Nevertheless, there are several
clearly identifiable types of decisions that are informed by testing, for which
some sort of classification system is useful. Because the actual use of a test
may change from what was originally planned, it is important to think in terms
of types of decisions more so than types of tests p er se; however,
it is common in actual usage to refer to types of tests as a convenient
shorthand.
0 komentar:
Posting Komentar