The phonological mediation hypothesis evidence from Chinese students with hearing impairment

Abstract

In Western countries, there are around 8-10% of school age children suffering

from mild to moderate reading difficulties. Similarly, in Hong Kong the prevalence

of reading difficulties among school age children was found to be 9.7- 12%. An

understanding of the mechanisms involved in reading helps us to determine the skills

necessary for successful reading.

The Phonological Mediation Hypothesis (PMH), which claims that

phonological recoding is a necessary phase during lexical access, is widely known

for its postulation that phonological awareness would be a significant prognostic

indicator of reading development. Theoretically, individuals who have difficulties in

phonological recoding during lexical access should also encounter reading

difficulties. Good readers with hearing impairment are, therefore, considered as

counter examples of PMH.

If the above prediction of PHM is true, skilled readers with hearing impairment

should have the ability to develop an intact phonological representation and hence

are more capable of phonological recoding. In this study, the reading behaviors of

children with hearing impairment (HI), that of their reading level matched (RL) and

that of their chronological age matched (CA) controls were compared in three

tasks—an auditory perceptual task of onset rime awareness (TAPOR); a synonym

decision task (SDT); and a homophone decision task (HDT). The results for TAPOR

showed that auditory perceptual ability (APOR) accounted for 49% of the variance in

the reading ability of children with hearing impairment. In addition, results of cross

group comparisons on the scores in TAPOR demonstrated a possible causal

relationship between APOR and reading ability in subjects with hearing impairment.

On the other hand, SDT and HDT results indicated a significant preference for

orthographic foils in RL and HI subjects with low reading ability. An increasing

tendency to choose synonyms or homophones, and a decreasing tendency to use

orthographic distractors was observed across subject groups with Primary 1, Primary

2 and Primary 6 reading abilities. A similar but delayed pattern of change in

preference for distractors was observed in HI subjects. The results only partially

agree with PMH. An alternative hypothesis—the Psycholinguistic Grain Size

Theory (PGST) — might be a better model to explain the observed results. With

reference to the results of TAPOR, the correlation between reading ability and

auditory perceptual ability could be explained in terms of the ‘availability problem’

postulated in this latter model. In the same way, the early emergence of orthographic

effect in almost all subjects except CA controls and the late emergence of a number

of reading strategies operating at different levels of grain size can be better explained

by ‘consistency problems’ and ‘granularity problems’ proposed in PGST. These

findings are considered, and directions for further studies are outlined.