Gary, a 10th grade biology teacher at Foothills High School, was shocked when he saw the results of his students' reading achievement tests. He knew that he had some low and high performers, but nothing could have prepared him for the enormous 15–grade-level spread in a single class with 24 students. One of his students was reading at a 3rd grade level; another had nearly hit the ceiling at 18.6. Within that tremendous range, nine students were reading at or below an 8th grade level, seven fell within a 9th–10th grade range, and eight students were reading at levels between 11th and 19th grade.When Gary learned that the readability level of his core textbook was 10.2, he realized that the textbook was essentially too difficult for more than one-third of his students to read. And for another third, the textbook might not be challenging enough to make the content interesting or engaging.
Unfortunately, countless high school teachers find themselves in Gary's situation. Without the benefit of a reading achievement test, they simply do not know their students' reading levels or how closely the instructional materials match their students' abilities.
Putting Testing in Its Place
Prominent critics have decried the current pervasiveness of standardized achievement testing in U.S. schools (see Kohn, 2000; Miller, 2001; Pearson, 2001; Popham, 2001). Respected literacy and education research organizations, including the International Reading Association (IRA), the National Council of Teachers of English (NCTE), and the American Educational Research Association (AERA), have issued carefully worded position statements warning against the abuses of high-stakes assessment. NCTE, for example,opposes the use of standardized, single-format testing as the sole or predominant determinant in decisions about students—their placement, progress, or program. (2000)
These critiques are, in our view, legitimate and necessary. In fact, we too have been critical of the overuse and abuse of standardized reading assessments (Brozo & Brozo, 1994; Brozo & Simpson, 2003; Hargis, 1999). In this highly charged atmosphere, however, many have come to regard achievement testing in all its forms as unwarranted and unnecessary, discrediting any sensible attempts to use achievement testing as a tool for teaching and learning.
Stiggins (2002) complains that standardized assessments of learning are increasingly being used to place blame, dole out punishments and rewards, and threaten students and teachers to increase effort. By contrast, he writes, assessment for learning is designed to help teachers craft more responsive curriculums that facilitate progress for all students. Using tests to determine students' reading ability levels and match them with the most appropriate instructional and recreational reading materials is more likely to promote learning than is using assessment to make placement and grouping decisions, establish a school or district performance rating, or index teacher merit pay to student achievement.
The best assessment for learning occurs within the context of daily instruction and situated literacy activities. The appropriate use of standardized reading achievement testing, however, can be a starting point to improved teaching and learning and can lead to much-needed reforms.
Low-Stakes Reading Achievement Testing
In September 2000, Foothills High School asked us to provide yearlong consultative services to fulfill the Tennessee Goals 2000 grant that the school had been awarded. The only high school in a small, four-school district, Foothills operates within a culturally and economically unique environment in east Tennessee. Situated in the shadow of a major corporation's aluminum plant tower, the school has been inextricably tied to the plant for more than a century. The recruitment of unskilled African American labor during the plant's boom years in the 1940s and 1950s has resulted in an ethnic diversity in Foothills and the surrounding community unlike any other in the region. The decline in U.S. metals manufacturing has led to an ever-growing unemployment rate. Nonetheless, the Foothills school district benefits from the plant's substantial tax support, which has made it possible to offer higher-than-average teacher salaries and build a new middle school and indoor sports complex.
The grant specified that we train and collaborate with teachers; share information about reading, conferences, and Web sites; evaluate and recommend library materials; acquire reading materials and assist in implementing them; and work with classroom teachers and individual students. The overarching goal of the grant, however, was to determine the reading abilities of all students and the effectiveness of initiatives to improve them.
To determine the range of student reading needs, we looked for reading achievement tests with recognized technical adequacy, sufficiently sensitive to measure achievement levels at the extremes of the reading achievement continuum in each of the high school grades. We also looked for assessments that were easy to administer to large groups so that we could minimize disruption to teachers' and students' daily routines. We decided on the Gates-MacGinitie Reading Tests (MacGinitie, MacGinitie, Maria, & Dreyer, 2000) and the Nelson-Denny Reading Test (Brown, Fishco, & Hanna, 1993). According to Buros (2001), both are psychometrically mature—in other words, both tests have been administered over a period of several years to large numbers of individuals and have been repeatedly analyzed to ensure technical soundness. The two in combination would enable us to find the most sensitive level of test difficulty for each student—particularly important because we hoped to use the instruments to measure end-of-year progress.
We expected to see a wide range of reading abilities in the Foothills students; therefore, we needed a means of determining which test or test level to assign to each student. Such an estimate of ability was vital to our plan to match individual students with the most sensitive Gates levels (which range as low as the equivalent of 1st grade) or with the Nelson-Denny test (with a ceiling achievement level of grade 18.9). We used STAR, the computerized reading test that accompanies the Accelerated Reader program, to determine a general instructional reading level for each student (see Buros, 2001, for a review of STAR's technical qualities).
With STAR results in hand, we assigned an appropriate Gates or Nelson-Denny level to students. If, for example, a student's STAR score indicated that his instructional reading level was between grades 7 and 8, he was assigned the Gates 7–9 form. If another student's STAR level was 12th grade, she was given the Nelson-Denny because its range includes the 11th grade level on the low end.
Teachers helped test all 346 students over the course of several days. After we scored students' Gates and Nelson-Denny pretests, we assembled the data and apprised teachers of the reading skill level of each of their students.
From Results to Reforms
Allington (2002) asserts that good teachers know their students well—they know where to begin appropriate instruction and how to recognize growth. An instructional starting point requires relevant information about students' skills, abilities, and knowledge. Without this essential information, teachers cannot account for individual differences in their planning or track students' progress.
Our initial meetings with Gary and his Foothills colleagues had indicated that they were not aware of the reading ability levels of their students. Now, faced with the pretest results, many Foothills teachers found themselves forced to adjust their assumptions about individual student performance in their classes. Martin, a math teacher, had to rethink whether 11th grader Tony's poor performance stemmed from “laziness,” as Martin had thought, or from the fact that he was reading at a 5th grade level. Pam, who taught sophomore English, began to wonder whether Tamika, who fell just one year shy of reaching the maximum on the Nelson-Denny by achieving a grade equivalent of 17.9, was disengaged in Pam's class simply because she was bored.
Overall, pretest results revealed that nearly 35 percent of Foothills students were reading at one or more grade levels below grade placement—and, in many cases, far below the expected grade level. Another 18 percent were reading at one or more grade levels above grade placement. With these results as a backdrop, we brainstormed several possible schoolwide literacy initiatives with department heads, the curriculum specialist, the principal, and his assistant. In the end, school administrators and a majority of the faculty endorsed three reform initiatives: sustained silent reading, reading young adult novels in the content classroom, and making alternatives to the textbook available for both struggling students and superior readers.
Sustained Silent Reading
Once students have mastered basic reading skills, the surest road to a richer vocabulary and expanded literacy is wide and sustained reading (Allington, 2002; Anderson, Wilson, & Fielding, 1988; Cipielewski & Stanovich, 1992; Taylor, Frye, & Maruyama, 1990). Yet many adolescents read less than their peers of 30 years ago (Carlson, 1999; Glenn, 1994; Libsch & Breslow, 1996) or, even more alarming, choose not to read at all (Beers, 1996; Schumm & Saumell, 1994). The less time young people spend with books and print, the less growth they exhibit on measures of vocabulary and reading achievement (Durrell, 1969; Glenn, 1994). This pattern seems to be particularly common among minority youth (Larson, Richards, Sims, & Dworkin, 2001), who score lower on achievement tests and are admitted to colleges in smaller numbers relative to other groups (Ogbu, 1994).
Advocates of secondary school reform initiatives agree that students need multiple opportunities for engaged, sustained print encounters in the classroom every day (Alvermann et al., 2002; Langer, 2000; Moore, Bean, Birdyshaw, & Rycik, 1999; Ruiz-de-Velasco, Fix, & Clewell, 2000). Lack of opportunities for regular, engaged reading may in part account for why most students who are poor readers entering high school remain at the low or basic reading level throughout their high school years (Cappella & Weinstein, 2001).
Foothills launched a sustained silent reading (SSR) program—or “Get Ready to Read,” as the students called it—to ensure that all students, regardless of ability, developed the reading habit and that all teachers participated in modeling the pleasure of self-selected reading. Using grant funds, the school purchased scores of high-interest young adult novels and magazines covering a wide range of reading levels, which were compiled into class sets. Staff members set up drop boxes throughout the building and encouraged students to donate a favorite paperback to the cause, netting many additional books for the classroom libraries. Teachers and students set aside one day each week during homeroom (approximately 25 minutes) for SSR—and many teachers added SSR days to their schedules as they discovered the advantages of focused, constructive, recreational reading for their students and for themselves.
Young Adult Novels
High school teachers rarely use novels anywhere except in the English classroom (Brozo & Simpson, 2003). Nevertheless, several Foothills teachers were willing to introduce novels into their classrooms once they realized that many of their students were unlikely to benefit from assigned textbook reading because the textbooks were too difficult for them. We also helped teachers overcome some initial reluctance by acquiring used and inexpensive class sets of books touching on a variety of content-area topics and themes and by conducting several classroom demonstrations with various novels.
Gary, the 10th grade biology teacher, used Farley Mowat's Never Cry Wolf (1979) in conjunction with a unit on food chains. During one class period, for instance, the day's lesson focused on population crashes. Gary defined the term, characterized the causes and effects of population crashes in the natural world, and asked students to find evidence and examples of such population crashes using the print and nonprint resources available to them. Working in pairs, many began their research by thumbing back through Never Cry Wolf to reference Mowat's description of the fate of the vole population when too many wolves were killed or relocated.
Following this student research, Gary organized a review of one chapter of the novel in “popcorn” format: He gave the class the first critical event of the chapter, then waited for students to stand up and take turns commenting on what happened next. Using this approach, the class covered most of the chapter plot in five minutes. Gary then asked for volunteers to come to the front of the room and take parts reading the next chapter aloud. After a few pages, a new group of students took over the parts and continued to read aloud. Gary was patient with all of his students and waited until one asked for help with a word before he or a classmate provided it. At several points in the lesson, Gary invited class members who were not reading aloud to ask questions of their classmates who were, which appeared to keep all students focused on the activity.
Alternative Texts
For the many Foothills students reading at two or more grade levels below their grade placement, engagement with the subject-area textbooks was unrealistic, if not impossible. Unfortunately, most high school teachers do not have at their immediate disposal texts in a range of difficulty levels on the same topic. The Internet, however, offers access to vast amounts of public domain material on virtually any school-related topic. In the area of chemistry, for example, we found easy readings on carbon bonding, the periodic table, and balancing equations, and we reformatted these texts to avoid stigmatizing struggling readers. The chemistry teacher then allowed his less skillful readers to use the alternative texts at the point during the 90-minute instructional block when students consulted their textbooks for information to solve problems or describe processes.
Janette, a U.S. government teacher, made numerous readings available from the Web on topics ranging from federalism to provisions of the Fourteenth Amendment. The level of difficulty of the texts was commensurate with the reading ability of her best readers, several of whom had scored on the Nelson-Denny test well above the 10.7 readability level of her course textbook. To ensure student engagement in the alternative readings, Janette had the high-ability students use them to gather additional information on textbook topics and report back to the class on what they had learned.
Posttesting
In mid-May 2001, just a few weeks before the conclusion of the school year, Foothills staff members readministered to students the tests that they had taken in November to determine whether students had made reading achievement gains. Posttest results indicated that nearly half of the students increased their scores by two or more grade levels, and another third maintained their pretest scores with an average grade equivalent level of 12.5. Seventy-three percent were reading at or above their grade placement, compared with 65 percent on the pretest measures.
Were these gains significantly greater than those that might have occurred without the literacy initiatives put in place at Foothills High School? We cannot say. This project did not involve a controlled experiment, nor were previous reading achievement data available to compare with our findings. We strongly believe, however, that the low-stakes reading achievement pretest results were the impetus for all of the reforms that staff members implemented. Once teachers and administrators became aware that perhaps as many as 20–25 percent of students were reading at two or more years below grade placement, they supported important literacy initiatives. Low-stakes reading achievement testing helped the Foothills teachers become much more sensitive to the importance of finding ways to accommodate the diverse reading needs of each of their students.
