The accurate perception of speech involves the processing of multidime
nsional information. The aim of this study was to determine the influe
nce of the semantic dimension on the processing of the auditory dimens
ion of speech by children with hearing impairment. The processing inte
ractions characterizing the semantic and auditory dimensions were asse
ssed with a pediatric auditory Stroop task. The subjects, 20 children
with hearing impairment and 60 children with normal hearing, were inst
ructed to attend selectively to the voice-gender of speech targets whi
le ignoring the semantic content. The type of target was manipulated t
o represent conflicting, neutral, and congruent relations between dime
nsions (e.g., the male voice saying ''Mommy,'' ''ice cream,'' or ''Dad
dy'' respectively). The normal-hearing listeners could not ignore the
irrelevant semantic content. Instead, reaction times were slower to th
e conflict targets (Stroop interference) and faster to the congruent t
argets (Stroop congruency). The subjects with hearing impairment showe
d prominent Stroop congruency, but minimal Stroop interference. Reduce
d Stroop interference was not associated with chronological age, a spe
ed-accuracy tradeoff, a non-neutral baseline, or relatively poorer dis
criminability of the word input. The present results suggest that the
voice-gender and semantic dimensions of speech were not processed inde
pendently by these children, either those with or those without hearin
g loss. However, the to-be-ignored semantic dimension exerted a less c
onsistent influence on the processing of the voice-gender dimension in
the presence of childhood hearing loss. The overall pattern of result
s suggests that speech processing by children with hearing impairment
is carried out in a less stimulus-bound manner.