Reading Between the Lines

When University of North Carolina system president Erskine Bowles commissioned some of his top researchers to conduct a study that compared education schools and their impact on the learning of K-12 students, his primary goal was to uncover the teaching methods that produce the best results.

And the study might have done exactly that, by uncovering a clear-cut choice between two reading methods.  Elementary reading education has been presumed to be a trouble spot since the publication of Why Johnny Can’t Read in 1955, and today literacy levels have been dropping in low-income communities and increasing numbers of college students now require remedial courses before tackling actual college-level material.

The study, conducted by UNC-Chapel Hill public policy professor Gary Henry and East Carolina University education professor Charles Thompson, examined North Carolina students’ incremental performances on standardized tests to see how much knowledge they gained in a specific year. They then matched the students with their teachers to determine the teachers’ effect.  Statistical techniques were used to remove the effect of such extraneous influences as students’ prior achievement levels, their family incomes, teachers’ pre-college preparation, and so on.

The results of the individual teachers were then aggregated according to how they entered the teaching profession—which education school they went to, whether they were part of the Teach for America program, or if they entered through the state’s own lateral entry program, and so on.

Because the study managed to determine the aggregate performance of graduates from specific education schools (or of teachers who took specific alternative paths of entry into the profession), the researchers were able to derive important conclusions. For instance, they discovered that Teach for America participants generally outperformed other teachers at the high school and middle school levels, sometimes by a wide margin. This suggests that something is amiss with the way teachers are recruited and taught at the state’s education schools, since Teach for America participants receive a scant six weeks of training before entering a classroom, while UNC-trained teachers receive a minimum of two years of training.

Another finding might prove to have an even greater impact than the discovery about Teach for America participants. Elementary teachers from UNC-Wilmington outperformed all others except for those from East Carolina University in both reading and mathematics.

But one education professor claims that it is not the reading methods favored by the school’s elementary education and literacy department that are producing such stellar results, but another method—one that he teaches in graduate classes, in an instructional design course and through extensive consulting in the area’s public schools.

The professor is Martin Kozloff, who teaches in the Educational Leadership Department at UNC-Wilmington’s education school. In an email exchange, Kozloff said that his teaching and consulting, and similar efforts by some like-minded colleagues, have likely skewed the study’s results, making UNC-Wilmington’s literacy program appear to outperform other UNC and private elementary education schools. This is because UNC-Wilmington’s graduates tend to teach in the local area in large numbers. Many of them have been trained in Kozloff’s methods in his classes and many others have been trained through his consulting (and that of his colleagues). His techniques differ from those espoused by the Watson School of Education at UNC-Wilmington.

Kozloff and his colleagues—UNC-Wilmington education professor John Rice and former New Hanover County director of instruction Frances Bessellieu—are  ardent proponents of a system of teaching known as Direct Instruction, an approach developed in the 1960s. In contrast, Bradford Walker, the chair of UNC-Wilmington’s Department of Elementary, Middle-level, and Literacy Education, said that his department favors a “whole language” approach to reading instruction.

According to Terry Stoops, a K-12 education analyst at the John Locke Foundation, DI and whole language approach reading instruction in very different ways. DI’s emphasis is on phonics—which teaches students to relate letters, the most fundamental building blocks of language, to their sounds. Whole-language instruction, on the other hand, teaches students to identify entire words and concentrates on their meaning.

Walker said in a phone conversation that “it is wrong to think that there is no phonics in whole-language instruction.” He said it is merely one part of the literacy puzzle—the others are syntax (how the language is put together) and semantics (meaning). “We think our students need to understand the whole process of reading, not just isolated processes,” he explained.

But Kozloff is dismissive of UNC-Wilmington’s method. He said that suggests that UNC-Wilmington trains prospective teachers to have their pupils “guess or predict what the words are using context clues such as pictures on a page.” He said that he and his colleagues “tested at least 2,500  ‘struggling readers’ in New Hanover County.” He said that they were trying to read “as they were taught.” And that was “by guessing.”

There is a large body of literature that supports Direct Instruction. For example, federal education agencies in the 1960s and 1970s found that direct instruction approaches reached “the greatest academic achievement gains among the students studied,” according to a 2007 study by the Pope Center.

Both Direct Instruction and whole language have a psychological component, but these differ greatly. Walker said that the UNC Wilmington education school tries to give its students a firm understanding of how a child learns to read—“what goes on in the mind of a child while they’re reading.”

Direct Instruction eschews such esoteric concepts, according to Stoops, and instead focuses on behavior modification.

So, while the actual reading instruction courses at UNC-W’s Watson School of Education are based on whole language, Kozloff and his like-minded colleagues have introduced DI into UNC-Wilmington’s education school and into the surrounding K-12 curriculum in several ways.

First, in the last ten years or so, Kozloff has taught over 1,000 students in EDN 301, Instructional Design, which is a required course for elementary education teachers.  In that course, he spends roughly two months teaching Direct Instruction methods.

He also said that he and Rice teach Direct Instruction to about 15-20 students per year who are getting their master’s degree in school administration. Many of these students go on to become elementary school administers in the local area, where they introduce DI to their faculties.

Kozloff also said that he, Rice, and Bessellieu conducted DI training in a large number of elementary schools in the three-county area surrounding Wilmington. In New Hanover County, Justine Lerch, the former director of elementary education, offered to pay for DI proprietary materials, and a publisher’s representative provided additional training. According to Kozloff, 20 out of 23 New Hanover schools adopted Direct Instruction to a “greater or lesser extent.”

The UNC-Wilmington graduates who have taken teaching jobs in the local area can easily be discovered, and it would be difficult for a researcher to discover which of those teachers received DI instruction from Kozloff, Rice, or Bessellieu, either through courses at the university or through in-house training at their schools. This makes the Wilmington area an excellent laboratory to uncover which of these two instructional philosophies is the more effective, using the same methods that Gary Henry and Charles Thompson employed in the original study.

Bowles has already expressed some interest in getting his researchers to more deeply analyze this situation. Perhaps, with his support, the original study will actually reveal a solution to a dire problem—unlike so many other studies that have come to naught.