Skip to Content

No 'Best Way' to Teach Reading


While the "reading wars" over the best way to teach reading have raged for decades, recent research suggests there may not even be a "best way."


Research published in the prestigious journal Science (Jan. 26, 2007) supports NEA's official position on the issue that sharply divides proponents of the phonics and supporters of the "whole language and meaning" approach and other teaching methods.


Researcher Carol M. Connor (Florida Center for Reading Research), an assistant professor in the FSU College of Education, and colleagues from FSU and the University of Michigan, wrote "Algorithm-Guided Individualized Reading Instruction." ( PDF, 184 KB, 2pp). The Science Daily web site reports that Connor's paper shows that lots of individualized instruction, combined with the use of diagnostic tools that help teachers match each child with the amounts and types of reading instruction that are most effective for him or her, is vastly preferable to the standard "one size fits all" approach to reading education that is prevalent in many American elementary schools.

"There is too much of a tendency in education to go with what 'sounds' really good," Connor said of various educational trends that come into and fall out of fashion. She told Science Daily, "What we haven't done very well is conduct comprehensive field trials and perform the rigorous research that are the norm in other fields of science. With this study, we sought to do just that -- to take a systematic approach to what works, what doesn't, and why" when teaching students to read."

The researchers' report itself explains:

"Much of the controversy regarding the best way to teach children how to read has focused on whether instruction should be code-based, such as phonics, or based on whole language and meaning, but this debate may miss the point. Although most children develop stronger reading skills when they receive a balance of explicit decoding instruction in combination with meaningful reading activities, even a balanced approach theory assumes that one approach, if it is the right one, will be equally effective for all children. Instead, the efficacy of any particular instructional practice may depend on the skill level of the student. Instructional strategies that help one student may be ineffective when applied to another student with different skills."

Connor said, "Instead of viewing the class as an organism, we're trying to get teachers to view the students as individuals."

Connor thinks technology can make that easier for classroom teachers to accomplish. She, Frederick J. Morrison and Barry Fishman, professors at the University of Michigan, have developed "Assessment to Instruction," or A2i, a Web-based software program. A2i uses students' vocabulary and reading scores and their desired reading outcome (i.e. their grade level by the end of first grade) to create algorithms that compute the recommended amounts and types of reading instruction for each child in the classroom. The software then groups students based on learning goals and allows teachers to regularly monitor their progress and make changes to individual curricula as needed.

A2i currently is being tested by about 60 elementary-school teachers in one Florida county. However, "right now A2i is just a research tool," Connor said. "Hopefully we'll be able to make it available more widely as time goes on."

In addition to Connor, Morrison and Fishman, other co-authors of the Science paper were Associate Professor Christopher Schatschneider of FSU's department of psychology and Phyllis Underwood, a doctoral student in the FSU College of Education.