Many schools-—especially ones in low-income areas—-have realized that there's a more efficient input towards the goal of maximizing testing performance than teaching academic subjects in a way that is rigorous and meaningful. This more efficient input is—test taking strategies!
Strategic test-taking involves a different set of actions than does reading and enjoying a chapter book—or reading, enjoying, and writing about a chapter book. For instance, on standardized reading tests at the 3-5 level, individual passages are about 500 words long, and there are about 8 or 9 questions per passage that total out at about 400-500 more words. It's more important to read the questions and answer choices carefully than the passage, and so it's come to be that teachers direct students to slowly read the questions first, then read the passage, underlining parts that may provide clues to answering the questions. Then, students are to again slowly ponder the answer choices while frequently referring back to the passage.
This read-questions-and-answers-then-scan-text-strategically approach isn't natural, but it works. Thing is, you can't introduce this strategy to students the week before The Big Test, or only a few will use it. You might be able to guess where I'm going here. To achieve high performance on standardized tests, it is perfectly sensible for teachers to have students read 500-word passages instead of chapter books all year long, and to read them in a way that will get them in the habit of strategically attacking multiple choice questions.
My radical suggestion is—-Don't test reading. For reasons that I don't have room to discuss here, I'm much more optimistic that critical thinking in math can be measured by the multiple choice format and that testing math doesn't lend itself to test score pollution in the same way that reading does. If every school in America administered the same rigorous math assessment for grades K-12, dataphiles at state education departments would have one incredibly useful measure of how well students are doing (by classroom, school, district, state, region, etc.). Creating such an assessment system, and eliminating the standardized test in reading, would promote the goal of meaningful accountability while delimiting that harm that strategic test preparation can do.
(For a longer version of this argument, see http://theedskeptic.blogspot.com/2010/03/dont-test-reading.html )