Low reading achievement for hearing impaired students has been an historic and consistent pattern. National studies from Pintner and Patterson (1916) to Allen (1986) have reported the average hearing impaired high school graduate reads at 4th grade level or lower. This study investigated the relationship of recall in short term memory, use of internal speech as a coding strategy, and reading comprehension. It compared reading achievement of hearing impaired students in Oral, Total Communication (TC), and Cued Speech (CS) communication modes. It replicated parts of the Conrad (1979) study, using his materials, procedures, and lists of acoustically similar and visually similar words.
The design utilized three communication modes (Oral, TC, CS) and two decibel levels (severe, profound); a Hearing group was used for comparison. Blocking factors were: decibel loss, age, general cognitive ability, years in communication mode, sex, and parent education level. Additional factors were: race, educational placement, communication support at home, and hand preference. 120 subjects, age 7-16, from 4 public school districts (which offer 3 communication mode tracks) were administered: the Raven SPM, 1982 SAT reading comprehension test, and the Conrad tests. The Conrad task was to read a series of one-word cards and write each word in order.
Results indicated hearing impaired students, as a group, attained significantly lower scores on the Raven, more errors on the Conrad test, lower SAT reading comprehension scaled scores, and lower IS ratios than did Hearing agemates. The TC group had significantly lower scores than Oral or CS. No significant differences were noted between: (1) decibel category groups or (2) reading achievement between Hearing and CS-profound groups as measured by SAT reading comprehension scaled scores. STM span was correlated with: (1) IS ratio in all hearing impaired groups and (2) reading in all four groups. Reading was correlated with IS ratio for O and TC, abut not for CS or Hearing.
Characteristics of the sample, coding strategies, and Type I and II coding are discussed. Implications for further research are presented.
