Wagner, R.K. & Torgesen, J.K., (1987). The nature of phonological processing and its causal role in the acquisition of reading skills. Psychological Bulletin, 101(2), 192-212.
My assigned article is Wagner and Torgesen’s (1987) literature review on the influence of phonological processing on reading ability acquisition. My outside reading is Allington and Woodside-Jiron’s (1999) critique of the uses and misuses of research in shaping education policy.
This article is coauthored by Joseph Torgesen and Richard Wagner, both of whom are currently at the Florida Center for Reading Research (FCRR) at FSU. They have been very influential in shaping reading policy in the state of Florida and nationally for the past several years. To add to my understanding of this article, I chose an article by Richard Allington, an outspoken critic of educational policy that has resulted from the use, or misuse, of research like Wagner and Torgesen’s. Allington’s piece is not a direct response to Wagner & Torgesen (1987), but it challenges the findings and their policy implications nonetheless.
In this literature review, Wagner and Torgesen wanted to see if there was a causal relationship between phonological processing and learning reading skills. They reviewed three distinct bodies of research that corresponded to three distinct aspects of phonological processing:
- phonological awareness (or phonemic awareness). This is the ability to perceive phonemes, measured by the ability to segment and blend phonemes in isolation.
- phonological recoding in lexical access. This refers to translating a written word into speech in order to access its meaning.
- phonetic recoding to maintain information in working memory. This refers to coding words into phonemes for efficient storage in working memory for processing.
The authors identify three major approaches to investigating the relationship between phonological processing and learning to read. The first is studying individuals with disabilities in reading and/or phonological processing. The authors did not consider these studies because they do not inform causal relationships. The two approaches considered for each phonological aspect were longitudinal correlational studies and experimental studies.
Wagner and Torgesen discuss the complications of trying to study questions of correlation and causation of such a complex and invisible process as reading. For example, two variables that seem related could both be influenced by a third variable that is not being measured. The researcher thinks she's looking at a cause and an effect, when she's only looking at two effects of the same unobserved cause. Likewise, there may be a web of unobserved causation at work in any study looking for correlation between variables, particularly when dealing with mental processes. Because direct observation of mental processes is not possible, a researcher may also confuse one variable with another, for instance unintentionally measuring working memory instead of, or in addition to, phonological ability. Likewise in experimental research, the researcher assumes that the training he provides to an experimental group is effectively targeting the skill or ability being studied. The degree to which this is true greatly affects the reliability of the results. The authors sought to offset the weaknesses of each type of study by combining studies of both types, longitudinal correlational studies and experimental studies, in their analysis.
Wagner and Torgesen find many areas of ambiguity and obscurity in the analysis of these data as it relates to their research question. They indicate areas within the reviewed studies where different methods would have answered a particular question, but the answer is unavailable because of the methods actually used. In every case, they describe what an ideal study to answer a particular question would look like, then describe the data from the studies that they found.
The authors found that the three aspects of phonological processing discussed in this article seem to be three ways of measuring the same underlying process. In addition, despite the methodological incompatibilities between Wagner and Toregesen’s research question and the data sets available, they reached the following conclusions:
The article that I selected by Allington & Woodside-Jiron was written more than a decade later in response to a widely-circulated white paper (Thirty Years of Research: What We Now Know About How Children Learn to Read) that purported to summarize best practices for classroom reading instruction. As the authors detailed, the research cited in the white paper did not adequately support the conclusions drawn. This white paper concerned these researchers because it strongly influenced educational policy in multiple states.
Although they don’t specifically address Wagner & Torgesen (1987), Allington and Woodside-Jiron criticize the types of conclusions drawn by Wagner and Torgesen. Specifically, the “studies have more often produced reliable, replicable gains for a specialized population only on measures of phonological processing and psuedo-word pronunciation tasks, while reliable, replicable gains on word reading, fluency, and prose comprehension have been more difficult to generate” (p. 3).
Allington & Woodside-Jiron offer the interpretation that studies of phonological processing provide evidence that: 1) 15-20% of students experience problems with phonological processing; 2) these problems are associated with early reading acquisition; and 3) difficulties in phonological awareness can be remedied. According to Allington & Woodside-Jyron, that’s it. Causation has not been established. There is insufficient evidence to make sweeping changes to instruction of entire school populations based on the research available. Research is often misappropriated in support of policy goals. These authors charge that the researchers involved benefit from having their research used as the basis for policy decisions and policy-makers benefit by adding the “stamp of approval” of scientific research to their policies. Simply put, “the use of 'research' as a policy advocacy tool seems less dependent on the reliability of synthesis of the research than on the ability to place 'research' summaries that support particular policy agendas into the hands of advocates and policymakers." (p. 11).
When reading research, I think it’s important to adopt a critical perspective not only toward sampling procedures and effect sizes, but toward the personalities and the politics connected to the research. In a perfect world, research would represent a dispassionate search for the truth. Considering the educational policy directions of the past twenty years, research like Wagner and Torgesen’s must be considered in a broader, and more critical, context.
References
Wagner, R.K. & Torgesen, J.K., (1987). The nature of phonological processing and its causal role in the acquisition of reading skills. Psychological Bulletin, 101(2), 192-212.
Allington, R., & Woodside-Jiron, H. (1999). The politics of literacy teaching: How "research" shaped educational policy. Educational Researcher, 28 (8), 4-13. Retrieved from: http://www.jstor.org/stable/1176311
Rather than a Swiss Army knife, filled with specialized tools for specific situations, Bates compares the language-enabled brain to a giraffe's neck -- adapted to serve a new function (reaching high leaves) while retaining its older functions (turning the head, passing air and food). Bates refers to the "Functional infrastructure for language" chart to show older adaptations combining to provide the basis for language. 
