Routh 1984; Olofsson and Lundberg 1983). Furthermore, evidence suggests
that training in these skills significantly improves early reading performance,
particularly in children with below average phoneme segmentation
skills (Bradley and Bryant 1985). (See Blachman 1984a, in press,
Wagner and Torgesen 1987, and Williams 1984 for a detailed review of
phoneme segmentation training studies.)
Two of the most promising training studies will be discussed here.
The first study, conducted in Sweden by Olofsson and Lundberg (1983),
provided segmentation training to groups of kindergarten children. The
results indicate that phoneme awareness skills can be developed in young
children. The game-like activities used in their study appear to be fun and
adaptable to diversified class groupings. However, the children who participated
in this study were a year older than the typical kindergarten
youngster in the United States. The effectiveness of training groups of
younger children to segment using these activities has not been
evaluated.
The second study, carried out in England by Bradley and Bryant
(1983), provided individual instruction for 5- through 7-year-old children.
Children were assigned to four experimenta ! groups: a) instruction in categorizing
words according to common sounds (such as by rhyme or alliteration);
b) instruction in categorizing words according to common
sounds and the use of plastic letters to represent those common sounds;
c) instruction in categorizing words according to semantic categories (control
group !); and d) no intervention (control group II). Bradley and Bryant
found that children who received sound categorization training outperformed
both control groups in reading, while those who participated in
sound categorization activities supplemented with alphabet letters were
the most successful group. This study offers the strongest evidence to
date of a possible causal link between phoneme awareness and reading
ability. However, the highly individualized nature of the training makes it
difficult to ascertain the potential of this type of instruction with groups of
children. Furthermore, one cannot tell from this study whether a group
that was exposed to letter names and letter sounds without sound categorization
training would have performed as well on reading and spelling
tasks as the groups that received sound categorization training.
The few training studies conducted in the United States provided
either one-to-one instruction or were conducted with older children (Hohn
and Ehri 1983; Wallach and Wallach 1976; Williams 1980). To date, no one
in the United States has evaluated a phoneme segmentation training program
with groups of kindergarten children as part of the regular school
day. Thus, the goals of our segmentation training project were as follows:
1. To replicate the finding that children in kindergarten can be taught