ABSTRACT
As with any endeavor in the area of educational measurement, assessing reading comprehension with a single test necessarily poses some limitations. Among the threats to reliability and validity in the use of a single instrument is the possibility that the content of the test does not fully represent the construct, and that the test taker is in some way disadvantaged due to the range of content that may be represented. The goal of the test developer is always to maximize the representativeness of the content sampled for the test, making it possible to generalize results to the content domain.
