The panel searched the literature to locate experimental studies that administered
systematic phonics instruction to one group of children and administered another type of
instruction involving either unsystematic phonics or no phonics to a control group.
Reviews were limited to experiments because these provide the strongest evidence that
instruction rather than some other factor caused the improvement in reading. Studies had
to examine phonics programs of the sort used in schools rather than used in laboratory
experiments focused on single processes.
One purpose of the meta-analysis was to determine whether there is evidence that
phonics instruction improves readers’ ability to read words in various ways.
Studies in the database were published between 1970 and 2000, although the majority
were conducted in the last 10 years: 1970-1979 (1 study); 1980- to 1989 (9 studies); and
1990 to 2000 (28 studies). Most (66%) were carried out in the United States, but 24%
were done in Canada, and the remainder in the United Kingdom, Australia, and New
Zealand. The primary statistic used to analyze effects of phonics instruction on outcome
measures was effect size, indicating whether and by how much performance of the
treatment group exceeded performance of the control group, with the difference
expressed in standard deviation units. The formula used to calculate raw effect size for
each treatment-control comparison consisted of the mean of the treatment group minus
the mean of the control group divided by a pooled standard deviation. This formula was
adopted for use in all meta-analyses conducted by the NRP.
Findings support the conclusion that systematic phonics helps children learn to
read more effectively than program with little or no phonics instruction. Phonics was
especially effective in teaching children to decode novel words, one of the main goals of