picture naming by children with hearing loss: i. effect … · j am acad audiol 13: 463-477 (2002)...

J Am Acad Audiol 13 : 463-477 (2002)

Picture Naming by Children with Hearing Loss : I . Effect of Semantically Related Auditory Distractors Susan Jerger*'' Lydia Lai*' Virginia A. Marchman*

Abstract

Thirty children with hearing loss (HL) and 129 typically developing (TD) children representing com-parable ages, vocabulary ability, or phonology skills named pictures while attempting to ignore semantically related or unrelated auditory distractors . The timing relation between the onsets of the distractors and pictures varied . A significant semantic interference effect, that is, slowed naming in the presence of the semantically related distractor, was observed in all groups, suggesting sim-ilar categorical knowledge in the HL and TD groups . The time course of semantic interference, however, was protracted in some children with HL, primarily those with unusually slow baseline naming speeds and early ages of identification/amplification of the loss . Thus, children with HL seem to develop normal lexical semantic representations . At the same time, the dynamics of semantic processing appear to be altered by the presence of early childhood HL .

Key Words : Childhood hearing loss, children, language processing, picture-word task, semantic interference

Abbreviations : DPDQ = Denver Prescreening Developmental Questionnaire ; HL = hearing loss ; HTL = hearing threshold level ; LSD = least significant difference ; NPVT = Rader Near Point Vision Test ; PPVT III = Peabody Picture Vocabulary Test-III ; SOA = stimulus onset asynchrony ; TD -typically developing ; VMI = Developmental Test of Visual Motor Integration ; VOR = voice-oper-ated relay ; WIPI = Word Intelligibility by Picture Identification test

Sumario

Treinta ninos con hipoacusia (HL) y 129 ninos en desarrollo tipico (TD) mostrando edades, vocabulario y habilidades fonol6gicas comparables, buscaron identificar dibujos tratando de ignorar elementos de distracci6n con o sin relaci6n semantica . La relaci6n temporal entre el inicio de los elementos de distracci6n y los dibujos vari6 . Un efecto de interferencia semantica significativo, esto es, una lentificaci6n en la identificaci6n en presencia de distracciones semanticamente relacionadas, fue observada en todos los grupos, sugiriendo un conocimiento categorizado similar tanto en el grupo con HL como en el TD . El curso temporal de la interferencia semantica, sin embargo, fue prolongado en algunos ninos con hipoacusia (HL), principalmente en aquellos con una velocidad de denominaci6n basal inusualmente lenta, y con edad tempranas en el diagn6stico/amplificaci6n de la perdida . Asi, los ninos con hipoacusia (HL) parecen desarrollar representaciones Idxico-semanticas normales . AI mismo tiempo, la dinamica del procesamiento semantico parece estar alterada por la presencia de una hipoacusia (HL) temprano en la infancia .

Palabras Clave : Hipoacusia infantil, ninos, procesamiento del lenguaje, tarea dibujo-palabra, interferencia semantica

Abreviaturas : DPDQ = Cuestionario Denver de Pre-Tamizaje en Desarrollo ; HL = hipoacusia ; HTL = nivel umbral auditivo ; LSD = diferencia menos significativa; NPVT = Prueba de Racer de Punto de Vision Cercana ; PPVT III = Prueba de Vocabulario por Dibujos de Peabody III ; SOA = asincronia del inicio del estimulo ; TD = en desarrollo tipico ; VMI = Prueba de Desarrollo de Integraci6n Viso-Motora ; VOR = dispositivo (relay) activado por la voz ; WIPI = Prueba de Inteligibilidad por identificaci6n de Dibujos

*School of Human Development, University of Texas at Dallas, Richardson : tThe Callier Center for Communication Disorders, Dallas, Texas

Reprint requests : Susan Jerger, School of Human Development, University of Texas at Richardson, PO Box 830688, GR4.1 . Richardson, TX 75083-0688

463

Journal of the American Academy of Audiology/Volume 13, Number 9, October 2002

W ords are the building blocks of spoken communication. To learn to commu-nicate, children must abstract and

store the meanings and names of words. The first stages of vocabulary development typically involve learning the meanings and names of com-mon objects, which requires the integration of cross-modal perceptual information . An important unresolved issue in the literature is how childhood hearing loss (HL) and its perceptual deficits affect the development of word representations and processes. Auditory input dominates word learn-ing in early childhood (Rufl`in-Simon, 1983 ; Tye-Murray, 1992 ; rye-Murray et al, 1995 ; Levelt et al, 1999). To the extent that HL degrades and fil-ters the auditory input, childhood HL may result in impoverished knowledge representations and slower, less efficient processing of linguistic information. Thus, studies investigating the nature of semantic knowledge in young chil-dren with a range of perceptual abilities should help to illuminate the role of auditory input in lexical learning .

The purpose of this research was to study the nature of semantic and phonologic process-ing in children with HL who use aural/oral com-munication . Two companion articles report the results, with this article focusing on the nature of semantic knowledge and the time course of semantic activation during word processing . We applied a newly developed children's picture-word task to assess the influence of semantically related auditory distractors on picture-naming speed (Jerger et al, 2002). This cross-modal par-adigm involves informational cross talk between the comprehension and production processing systems (Levelt et al, 1991). The patterns of this cross talk provide a basis for hypothesizing about the nature of semantic knowledge and the mechanisms underlying language use (Rayner and Springer, 1986; Glaser, 1992). We present a model of performance on the picture-word naming task, followed by predicted results in children with HL.

1999). The dependent measure is the speed of pic-ture naming. Typically, a categorically related distractor increases the latency of picture nam-ing relative to unrelated distractors or to another type of baseline, termed the semantic interference effect. Another important manipulation is that the onset of the distractor is varied to be before, after, or simultaneous with the onset of the picture, termed the stimulus onset asynchrony (SOA). As you will see, both the type of distractor and the SOA are critical determinants of whether a spo-ken word influences the speed of picture naming (e.g ., La Heij et al, 1990 ; Meyer and Schriefers, 1991).

Several theoretical models of performance on the picture-word naming task have been sug-gested . In general, the models propose that lexi-cal access is subdivided into semantically and phonologically constrained stages of processing. The relation between the components has been con-ceptualized as discrete, cascaded, or interactive, a debate that is not critical to the issues investi-gated here (for further information, see Dell, 1986; Dell and O'Seaghdha, 1991 ; Levelt et al, 1991 ; Roelofs,1992 ; Jescheniak and Schriefers,1998 ; Cut-ting and Ferreira, 1999 ; Damian and Martin, 1999).1 The model of Levelt and his colleagues (1991), depicted in Figure 1, ably conceptualizes our research. The model includes the following stages for picture naming : (1) conceptual pro-cessing, (2) activation of a set of meaning-related lexical items, (3) output phonologic processing of the selected item, and (4) naming response . In con-trast, processing of the auditory distractor is assumed to consist of the following stages : (1) auditory/phonetic processing, (2) input phono-logic processing with activation of a set of phono-logically related lexical items, (3) lexical-semantic processing, and (4) conceptual processing.

Figure 1 illustrates the temporal relations, as hypothesized by Levelt and colleagues (1991), among the assumed stages of processing for two

Cross-Modal Picture-Naming Task: Semantic Condition

The picture-word task is a double-stimulation paradigm. A participant is asked to name a pic-ture while attempting to ignore an auditory dis-tractor. In the semantic condition, the distractor is either categorically related or unrelated to the picture, for example, the picture "apple" coupled with "peach" (related, i.e ., food category) and "nickel" (unrelated) (from Damian and Martin,

'In contrast to the discrete two-stage model, the cas-cade and interactive-activation models propose that all activated lexical items, not just the selected item, become phonologically encoded and that the time courses of seman-tic/syntactic and phonologic activation overlap . Even an interactive model such as that of Dell and O'Seaghdha (1991), however, assumes that there is greater activation at the lexical-semantic level than at the phonologic level early in the time course of picture naming, and that at some later point, after the lexical-semantic representation or node with the highest activation is selected and given a jolt of activation, there is a much higher degree of activation for the phonologic representation of the selected word .

464

Semantically Related Disorders/Jerger et al

Input Auditory/ I ~ Phonetic

Input Phonology

Onset Auditory -o Di stractor

SOA : -150 Input Visual Processing

Picture Onset

Onset `No- Auditory

Distractor

SOA: +150

Input Phonology

i

Output Phonology

Lexical Semantic

Output Phonology

i Naming Response

Concept

Figure l Temporal relations, as hypothesized by Levelt et al (1991), among the assumed stages of processing for two

stimulus onset asynchronies (SOAs) : -150 msec (the auditory distractor leads the picture) and +150 msec (the audi-tory distractor lags the picture) . The shaded areas represent temporal overlap between the stages of picture-distrac-tor processing . Redrawn from Jerger et al, 2002b.

SOAs: -150 msec (the auditory distractor leads the picture) and +150 msec (the auditory dis-tractor lags the picture) . The schematic provides a possible explanation for the finding that normal adult and child performance on the picture-word task yields a pattern of semantic interference at -150 msec SOA but not at +150 msec SOA (Schriefers et al, 1990 ; Damian and Martin, 1999 ; Jerger et al, 2002b) . In other words, the interfering effect of a semantic distractor on picture naming is greater when the distractor is presented before the onset of the picture (-150 msec SOA). Seman-tic interference is hypothesized to occur when the picture-naming process is occupied with lex-ical retrieval. A slowed picture-naming response occurs because the distractor's and the picture's lexical semantic representations are coactivated within the same time window. The idea is that the semantically related distractor and the picture share common semantic features as members of the same superordinate category (see Glaser, 1992, for discussion) . For example, the picture-distractor pair "dog" and "cat" are both members of the animal category, sharing seman-tic features such as pet, furry, four legs, etc.

Semantic interference occurs because coactivated, meaning-related lexical representations compete for selection and control of the naming response (the lexical competition hypothesis, see Damian et al, 2001, for a discussion) .

In short, the basis of semantic interference is temporal overlap between the picture's and the distractor's lexical semantic representa-tions, which is greater at -150 msec than at +150 msec SOA, as depicted in Figure 1 . When the categorically related distractor lags the onset of the picture (+150 msec SOA), the pic-ture's word entry is assumed to be selected prior to the distractor's complete lexical seman-tic activation, and no interference is observed . A question is whether this normal pattern of results will be observed in children with mod-erate HL, which is hypothesized to limit the abstraction and storage of some word mean-ings in semantic memory.

Predicted Results in Children with HL

Semantic interference is determined by both word meaning and categorical knowledge.

465


With regard to word meaning, children with a moderate degree of HL show a reasonably nor-mal pattern of vocabulary development, although the rate is slowed and individual vari-ability is pronounced (Blair et al, 1985 ; Osberger and Hesketh, 1988 ; Gilbertson and Kamhi, 1995 ; Briscoe et al, 2001). Word-learn-ing ability, and measures of language ability in general, is not readily predicted from the degree of HL. This counterintuitive finding may be related to important differences in factors such as the age of onset of the loss (Davis et al, 1986 ; Gilbertson and Kamhi, 1995). With regard to the organization of semantic categories in children with HL, categorical knowledge appears normal for categories that are easily perceived visually (Osberger and Hesketh, 1988) . Individuals with HL also show normal proactive interference when recalling categor-ically related pictures and normal release from proactive interference when the final to-be-recalled picture is changed to a different seman-tic category (Hoemann et al, 1974). Individuals with even severe to profound HL can make efficient use of semantic relations to boost per-formance on a paired associate memory test, learning "fork-knife" better than "bell-leaf' (Campbell and Wright, 1990). Finally, children with a mild to moderate degree of HL demon-strate the auditory Stroop effect, another form of semantic interference (Jerger et al, 1988, 1997). Reaction times to talker gender are con-sistently slower and less accurate for speech tar-gets with a semantically conflicting relation between stimulus dimensions (e.g ., a male talker saying "Mommy") than with a semanti-cally congruent dimensional relation (e.g ., a male talker saying "Daddy") .

This review overall predicts semantic inter-ference effects on the picture-word naming task in children with HL owing to the reasonably normal semantic knowledge demonstrated in this population . For this study, we ensured the familiarity of picture-word pairs by selecting words that represented an early age of acquisi-tion, high familiarity, high concreteness, and high imageability (Paivio et al, 1968 ; Carroll and White, 1973a, 1973b; Gilhooly and Logie, 1980 ; Snodgrass and Vanderwart, 1980; Fran-cis and Kucera, 1982; Nusbaum et al, 1984 ; MacWhinney, 1993). With regard to the dynam-ics of the information processing system in chil-dren with HL, predictions are difficult to derive from the currently available literature . Our results will provide new information about the nature of semantic representations and the time

course of lexical semantic processing in chil-dren with HL.

METHOD

Participants and Demographics

General

The participants, who were recruited from cooperating educational programs, were divided into a group with HL and several typically devel-oping (TD) comparison groups . The criteria for participation were English as the native lan-guage and no diagnosed or suspected disabilities (excepting, in the HL group, the speech and lan-guage problems that accompany childhood HL) . All of the children had normal visual acuity, gross neurodevelopmental history, visual-motor integration, and oral-motor function . The aver-age Hollingshead Social Strata Score (1.6) was consistent with a major business and profes-sional socioeconomic status .

HL Group

The group consisted of 16 boys and 14 girls between 5 and 15 years of age. The racial/eth-nic distribution was 74 percent Whites, 13 per-cent Hispanics, and 13 percent Blacks . All of the children were considered successful hearing aid users and attended regular classes (main-streamed), with support from a speech-language pathologist or special education teacher. Table 1 details the audiologic characteristics . The par-ticipants were tested while wearing their hear-ing aids . A technician routinely inserted new batteries into the aids and ensured that they were functioning properly prior to testing.

Comparison Groups

One hundred twenty-nine TD children were divided into groups representing comparable ages, receptive vocabulary skills, or input phono-logic skills . The racial/ethnic distribution was 80 percent Whites, 8 percent Asians, 7 percent His-panics, and 5 percent Blacks . Hearing sensitiv-ity, speech detection thresholds, and word recognition scores were within normal limits in all children. The age comparison group consisted of 60 children (n = 30 boys and 30 girls, 5 to 14 years of age) . The receptive vocabulary compar-ison group contained 60 children (n = 35 boys and 25 girls, 4 to 12 years old) . The input phonology comparison group consisted of 30 children

466


Table 1 Audiologic Characteristics of the Participants with Hearing Impairment

Metric

Age at suspected onset of hearing loss

Results (SD)

Unknown : 6 Congenital/at birth : 16 Birth to 2 yr 7 2 through 3 yr : 1

Age at initial identification/amplification of hearing loss

Etiology of hearing loss

Unaided hearing sensitivity Better ear, PTA Poorer ear, PTA Better ear, 4000 Hz Poorer ear, 4000 Hz

Aided performance Speech detection threshold Word recognition

Prior to 2 yr : 7 2 through 3 yr : 11 4 yr : 9 5 through 7 yr : 3

Unknown: 13 Hereditary/genetic : 8 Meningitis : 2 Birth problems : 2 Trauma : 4 Ototoxicity: 1

50.56 dB HTL (20 .59) 60.96 dB HTL (21 .94) 57.76 dB HTL (25 .48) 70.34 dB HTL (24 .67)

13.79 dB HTL (7 .40) 89.60% (16.56)

Unaided hearing sensitivity was summarized by averaging hearing thresholds levels (HTLs) across 500, 1000, and 2000 Hz (pure-tone average [PTA]) .

(n = 20 boys and 10 girls, 3 to 6 years old). Eleven children were in both the vocabulary and phonol-ogy groups and 10 were in both the vocabulary and age groups. There were no overlapping mem-bers in the phonology and age groups .

Demographic Materials and Instrumentation

Hearing sensitivity was assessed with a standard pure-tone audiometer. Word recog-nition was assessed with the Word Intelligibility by Picture Identification (WIPI) test (Ross and Lerman, 1971). The test items were recorded by a male talker with general American dialect into a computer and played back via a speech audiometer and associated loudspeaker. The sampling rate was 22 kHz with 16-bit ampli-tude resolution . The output intensity levels of the stimuli were adjusted to equivalent peak intensities .

Visual acuity was screened with the Rader Near Point Vision Test (NPVT) (Rader, 1977). Oral-motor function was screened with a ques-tionnaire designed by an otolaryngologist who is also a speech pathologist (Peltzer, 1997). Gross

neurodevelopmental deficits were ruled out with the Denver Prescreening Developmental Ques-tionnaire (DPDQ) (Frankenburg et al, 1976) in the younger participants and with the medical and educational histories in the older partici-pants with normal hearing and in all of the par-ticipants with HL. Socioeconomic status was estimated with the Hollingshead Four-Factor Index (Hollingshead, 1975).

To estimate nonverbal abilities, visual per-ception was assessed with the Southern Cali-fornia Figure Ground Visual Perception Test (Ayres, 1978) in children 8 years of age or younger and with the Block Design subtest of the Wechsler Intelligence Scale for Children-Revised (WISC-R; Wechsler, 1974) in children 9 to 15 years. Nonverbal reaction time was quantified as the average (10 trials) time to respond with a key press to a visually presented neutral stim-ulus . Visual-motor skills were estimated with the Developmental Test of Visual Motor Integra-tion (VMI) (Beery, 1989) .

To estimate verbal skills, vocabulary knowl-edge was assessed with the Peabody Picture Vocabulary Test-III (PPVT III) (Dunn and Dunn, 1997). Input phonologic knowledge was esti-

467


mated by onset and rhyme laboratory mea-sures, as detailed in the companion paper (Jerger et al, 2002a) . Output phonologic knowl-edge was estimated by articulatory proficiency (Goldman-Fristoe Test of Articulation ; Gold-man and Fristoe, 1986). Finally, mixed seman-tic/phonologic ability was estimated with the Expressive Vocabulary Test (EVT) (Williams, 1997) .

Demographic Procedure

Suellen acuity (including participants with corrected vision)

4. Gross neurodevelopmental status-within the pass criteria for the child's age on the DPDQ or on the medical and educational histories

5. Visual-motor integration-score above the tenth percentile on the VMI

6. Oral-motor function-no difficulty with any of the five items on the Peltzer questionnaire concerning eating, swallowing, and drooling

The battery of demographic measures was administered randomly, with three exceptions . The measures using items from the picture-word test were administered after completing the picture-word testing . The expressive vocabu-lary and articulation tests were administered prior to the receptive vocabulary test . The VMI test was always the first test because it seemed a helpful tool for establishing rapport. Each of the standardized measures was administered and scored according to the recommended tech-nique . The onset-segment and rhyme measures were administered as described in the compan-ion article (Jerger et al, 2002a) .

The operational definitions for normalcy were the following:

1. Hearing sensitivity-bilaterally symmetric hearing threshold levels (HTLs) (ANSI, 1989) of <_ 20 dB at all test frequencies between 500 and 4000 Hz on both ears

2. Speech recognition ability-score of at least 90 percent correct on the WIPI test .

3. Visual acuity-100 percent correct identifi-cation of targets on the NPVT at 20/25

Comparison of Groups

Table 2 summarizes results on the vari-ables used to equate performance in the HL and comparison groups . Statistical analyses were carried out to determine whether performance on the equated variables was equivalent between groups. All analyses were negative, indicating no significant differences (all p values > .76) . Results for the HL and age comparison groups showed that chronologic ages and nonverbal skills were equivalent between groups . Results for the HL and receptive vocabulary comparison groups showed that vocabulary knowledge was comparable between groups . Results for the HL and input phonology comparison groups showed that input phonologic skills for audiovisual stim-uli were comparable between groups .

With regard to vocabulary skill in the groups representing other comparisons, receptive vocab-ulary ability differed significantly between the HL and age groups and the HL and input phonol-ogy groups . The HL group had a lower PPVT III raw score and percentile score than the age com-parison group, with raw scores of 116 and 150,

Table 2 Performance on Variables Equated between HL and Comparison Groups (Ranges in Parentheses)

Hearing Impairment (n = 30)

Age Comparison (n = 60)

Chronologic age (mo) 128 (64-185) 125 (64-177) Visual perception (age equivalency, mo) 145 (66-203) 149 (69-203) Visual simple reaction time (msec) 504(287-728) 513(287-728)

Receptive Vocabulary Comparison (n = 60)

Receptive vocabulary (raw score) 116 (38-197) 115 (51-190)

Input Phonology Comparison (n = 30)

Onset (%) 96 (90-100) 99 (90-100) Rhyme (%) 91 (60-100) 95 (60-100)

respectively, t = -4.95, p = .0001, and percentile scores of 31st and 81st, respectively, t = -9.47, p = .0001 . The HL group had a higher PPVT III raw score, but lower percentile score, than the input phonology group, with raw scores of 116 and 88, respectively, t = 3.56, p = .0008, and per-centile scores of 31st and 76th respectively, t = -6.96, p = .0001 . With regard to age in the groups representing other comparisons, age dif-fered significantly between the HL and input phonology groups and the HL and vocabulary groups . The HL group was about 51/a years older than the input phonology group, t = 10.78, p = .0001, and about 4 years older than the vocabulary group, t = 8.03, p = .0001 .

Experimental Tasks

InstrumentationlMaterials

For the picture-word task, colored pictures were scanned into a computer as 8-bit PICT files and edited to achieve objects of a similar size and complexity on a white background . The auditory distractors were recorded directly into the computer by the same male talker as above. The sampling rate was 22 kHz with 16-bit amplitude resolution . The pictures were pre-sented via a computer monitor. The auditory distractors were presented via the speech audiometer and associated loudspeaker. The output intensity levels of the stimuli were adjusted to equivalent peak intensities . Both the computer monitor and the loudspeaker were mounted on an adjustable height table directly in front of the child at a distance of approximately 90 cm . When naming each pic-ture, the child spoke into a directional micro-phone mounted on an adjustable stand. To obtain naming latency, the computer triggered a counter/timer with 1 msec resolution at the initiation of a target . The timing board was stopped by the onset of the child's vocal response into the microphone, which was fed to a voice-operated relay (VOR). A pulse from the VOR stopped the timing board, which displayed the time in fractional seconds.

The development of the items and details regarding the Children's Picture-Word Test has been detailed previously (Jerger et al, 2002b) . In brief, children were asked to name pictures in the presence of various types of auditory distractors that represented either semantic or phonologic relations between the pictures and the words (see also Jerger et al, 2002a) . For the semantic items, naming was evaluated in the presence of


four different types of auditory distractors, rep-resenting a categorically related word, an unre-lated word, a verbal baseline, and a nonverbal baseline . For example, when asked to name the picture pizza, auditory distractors were the related word "hotdog," the unrelated word "dress," and the baselines "five" and a drumroll sound . The drumroll sound was the average duration of the other distractors (515 msec). Eight pictures were named in the presence of each type of dis-tractor, for a total of 32 trials .

In addition to the picture-word task, two additional tasks assessed the effectiveness of the distractors for each individual child. First, we assessed whether each child appreciated the categorical relation between the picture-word pairs. The eight semantically related picture-word pairs were displayed on a picture-response card containing four other pictures, that is, the categorically related pair (knife-gun) plus foils (cheese, dog, fish, present) . Different response cards varied the position of the alternatives across children . Second, we assessed whether each child could perceive the 16 auditory word distractors and the verbal baseline accurately. Each child repeated each of the words as they were presented over the loudspeaker. Then the experimenter asked the child to identify each word on a picture card . Each card had six pic-tures, the target and five foils with the same number of syllables as the target. Different response cards varied the position of the alter-natives across children .

Procedure

Testing was carried out within a 12 x 14 ft double-walled sound-treated booth. The par-ticipant was seated at a child-sized table, and the computer monitor was adjusted to be at eye level. A tester sat at the computer console to administer the experiment, and a cotester sat alongside the child, keeping him/her focused on the task . All trials judged to be flawed were deleted on-line and readministered after inter-vening items, for example, lapses of attention, squirming out of position, triggering the micro-phone in a flawed manner, and mis-hitting the response key. The intensity level of the auditory stimuli was approximately 70 dB SPL, as mea-sured at the imagined center of the partici-pant's head with a sound level meter. The child's responses to the picture-word test, the stan-dardized articulation test, and the laboratory repetition measures were digitally recorded and scored off-line by two speech pathology stu-

469


dents. The participants received a small toy for participating.

Prior to beginning the picture-word task, a simple naming task without any auditory dis-tractors was completed. The target picture was always "lamp." The task quantified the child's ability to detect and name a predetermined tar-get. The simple naming task was repeated at the completion of all picture-word testing. Each pre-sentation consisted of 2 practice trials and 10 test trials .

For the picture-word task, the child was instructed to name each picture as quickly and as accurately as possible and to ignore the audi-tory distractor. The picture remained in view until the child's naming response . Each trial was initiated by the tester's pushing the space bar, out of sight of the child. The picture-word task was administered in a practice session prior to the primary experimental procedure. During the practice run, each picture was pre-sented once, coupled with one of its auditory distractors selected at random . The SOA was held constant at 0 msec . At the end of the prac-tice trials, the tester showed each picture on a 5 x 5" card to the child, teaching her the target names of any pictures named incorrectly and helping her practice saying the names without "a."2 During the primary run, each picture was presented with each of its auditory distractors at each of three SOAs : -150, 0, and +150 msec. The test items and the SOAs were presented ran-domly within one unblocked condition (see Star-reveld and La Heij, 1996, for a discussion) . No picture or distractor was allowed to recur with-out at least one intervening item .

For the category knowledge measure, a tester spoke the nominal descriptor for the tar-get category for each of the semantic picture-word pairs ("Which ones are food?"). The child responded by pointing to the two correct pictures on the response card (pizza-hotdog) . If the child responded incorrectly, the tester spoke a func-tional descriptor for the category ("Which ones do we eat?") . The child's score was percentage cor-rect performance. For the distractor perception task, each distractor was presented auditorily, and the child was asked to repeat it and then to identify it from a set of pictured alternatives . The

'Children's tendency to say "a" before a picture's name is a significant problem . Instructing them not to do this makes the tendency worse . Practice naming runs, alter-nating between the tester and the child, with the tester illustrating the correct style, seemed particularly valuable .

scores were percentage correct repetition and identification . Consistent mispronunciations were not scored as incorrect.

Data Analysis

The total number of deletions (with replace-ment) represented 9.6 percent (HL), 7.0 percent (age), 7.8 percent (vocabulary), and 9.9 percent (phonology) across all trials . The total number of missing trials after replacement was 2.1 per-cent (HL), 0.7 percent (age), 1 .2 percent (vocab-ulary), and 1.7 percent (phonology) of overall trials . Naming responses that were more than three standard deviations from an item's con-ditional mean were also discarded. This proce-dure resulted in the exclusion of 0.6 percent (HL), 0.4 percent (age and vocabulary), and 1.7 percent (phonology) of trials . Overall, a total of 2 .7 percent (HL), 1.1 percent (age), 1.6 percent (vocabulary), and 3.4 percent (phonology) of tri-als were missing or excluded. The remaining picture-word results were analyzed by subjects (Fl) and by items (F2) with a mixed-design analy-sis of variance . The F2 analyses have reduced sta-tistical power relative to the Fl analyses owing to the small number of items in each condition (n = 8) .

Owing to the overlapping membership in the various TD groups, three separate analyses were carried out: HL versus age, HL versus vocabulary, and HL versus input phonology. Individual variability in the degree of interac-tion between the picture-distractor pairs was fur-ther assessed with multiple regression analysis in the HL group. To address a multicollinearity problem among some of the demographic vari-ables (e .g ., significant correlations between age, PPVT III, etc.), data were transformed into stan-dard scores, and a single composite variable was formed to represent age/age-related vari-ables (Pedhazur, 1982).

RESULTS

Baseline Condition

Figure 2 shows average naming latencies in the groups for the verbal and nonverbal baseline distractors. Average overall naming latencies ranged from about 865 msec in the age com-parison group to 1445 msec in the input phonol-ogy comparison group. The slower naming times in the phonology and vocabulary groups, relative to the HL and age groups, are consistent with their younger ages . Age-linked improvements in

470


Phonology Comparison

1750

1500

1250

1000

750

Age Comparison

8

v

on a

ca z

1750

1500

1250

1000

750

Hearing Impairment

-150 0 150 -150

El Word "Five"

" Drumroll

Receptive Vocabulary Comparison

0 150 -150 0 150

Stimulus Onset Asynchrony (msec)

Figure 2 Average naming latencies in all groups in the presence of the baseline distractors, the syllable "five" and the drumroll sound . The height of the symbol obscures the error bars for some data .

reaction time have been reported for decades (Goodenough, 1935 ; Jerger et al, 1988). Three statistical analyses by subjects and by items included the between factor of group (HL vs age, HL vs vocabulary, or HL vs input phonol-ogy) . Each analysis contained the same within factors, namely SOA (-150, 0, and +150 msec) and distractor type (word and drumroll). Over-all naming latencies changed significantly as the SOAvaried from-150 to +150 msec : HL-age : F1 = 33.28, df = 2, 176, p = .0001; F2 = 31.93, df = 2, 28, p = .0001; HL-vocabulary: F1= 43.07, df = 2, 176, p = .0001; F2 = 49 .52, df = 2, 28, p = .0001; HL-phonology : F1= 22.37, df = 2, 116, p = .0001; F2 = 35.38, df = 2, 28, p = .0001. Nam-ing responses were relatively delayed when the distractor's onset was after the picture's onset (+150 msec SOA) . The variation in naming times across SOA was greater in the vocabulary and phonology groups than in the HL group, with a significant SOA x group interaction: HL-vocabulary : F1 = 6.45, df = 2, 176, p = .002 ; F2 = 9.09, df = 2, 28, p = .0009; HL-phonology : F, = 4.91, df = 2, 116, p = .009 ; F2 = 10.37, df = 2, 28, p = .0004.

The type of distractor also significantly affected the baseline naming latencies : HL-age : Fl = 32.32, df = 1, 88, p = .0001 ; F2 = 38.70, df = 1, 14, p = .0001 ; HL-vocabulary : F1= 38.06, df = 1, 88, p = .0001 ; F2 = 46.17, df = 1, 14, p = .0001 ; HL-phonology : Fl = 22.53, df = 1, 58, p = .0001 ; F2 = 40.88, df = 1, 14, p = .0001 . Over-all naming times in all groups were slower in the presence of the word distractor than the drumroll distractor. This finding agrees with previous results in adults (Martin et al, 1988 ; LeCompte et al, 1997) . The type of distractor influenced naming more in the vocabulary and phonology groups than in the HL group, with a significant type x group interaction, HL-vocabulary : F1 = 5 .70, df = 1, 88, p = .02 ; F2 = 6.75, df = 1, 14, p = .02 ; HL-phonology: F1 = 3.70, df = 1, 58, p = .05 ; F2 = 10 .37, df = 1, 14, p = .006 . In the results below, we subtracted each participant's baseline from his/her exper-imental measures to render the experimental effects easier to discern graphically. The patterns of semantic interference were comparable for the different baselines . Thus, data are presented only for the verbal baseline .

471


Semantic Interference Effect

Prior to analyzing results, the individual data were refined on the basis of performance on the category knowledge and distractor recog-nition tests. Results showed that all of the chil-dren appreciated the categorical relation between all of the pictures and their related auditory distractors. No modifications to the individual data were indicated. With regard to the distractor recognition test, all TD children and 24 HL children correctly identified all of the semantically related distractors . Six of the HL children misheard from one to three related distractors, one (n = 3), two (n = 1), three (n = 2) . Each picture-word pair containing a misheard distractor was deleted from the indi-vidual data . Performance in these six children was, thus, the average of seven (misheard one), six, or five picture-word pairs rather than eight pairs. With regard to the unrelated distractors, all of the TD children and 21 of the HL children correctly identified all items. Nine of the HL chil-dren misheard from one to four distractors, one (n = 5), two (n = 3), four (n = 1) . Since each mis-heard word was also semantically unrelated to

V V y

v C d ca a 00 a m z d y

a

d

U

400

300

200

100

0

400

300

200

100

0

the picture, we did not eliminate any items from the individual data of these nine HL chil-dren . Statistical analysis indicated that results for the semantic condition did not differ in the HL children with 100 percent correct versus less than 100 percent correct distractor recog-nition . The data below for the HL group, there-fore, represent results in all 30 children .

Figure 3 shows average adjusted naming latencies for the related and unrelated distrac-tors in all groups . Both types of distractors pro-duced some interference in performance in all groups relative to the baseline . This finding agrees with previous findings in adults (Schriefers et al, 1990). Statistical analyses by subjects and by items included the between fac-tor of group (HL and normal) and the within fac-tor of SOA and type (related and unrelated) . Overall adjusted naming times declined signif-icantly in all groups as the SOA varied from -150 msec to +150 msec : HL-age : Fl = 5.71, df = 2, 176, p = .004 ; F2 = 8.64, df = 2, 28, p = .001 ; HL-vocabulary: Fl = 6.41, df = 2, 176, p = .002 ; F2 = 7.03, df = 2, 28, p = .003; HL-phonology : Fl = 4.29, df = 2, 116, p = .02 ; F2 = 4.93, df = 2, 28, p = .01 . More importantly, the type of

Phonology Comparison

Hearing Impairment

O Related " Unrelated

Receptive Vocabulary Comparison

-150 0 150 -150 0 150 -150 0 150

Stimulus Onset Asynchrony (msec) Stimulus Onset Asynchrony (msec) Stimulus Onset Asynchrony (msec)

Age Comparison

Figure 3 Average adjusted naming latencies (experimental minus baseline latencies) in all groups in the presence of semantically related and unrelated word distractors . The baseline was the syllable "five ."

472


distractor significantly affected performance : HL-age : F, = 62 .97, df = 1, 88, p = .0001 ; F2 = 17.56, df = 1, 14, p = .0009 ; HL-vocabulary : F, = 35.10, df = 1, 88, p = .0001 ; F2 = 18.33, df = 1, 14, p = .0008 ; HL-phonology : F, = 16.83, df = 1, 58, p = .0001; F2 = 3.84, df = 1, 14, p = .07 . Naming performance in all groups was slower in the presence of the semantically related distractors . The difference in naming latencies for the two types of distractors, that is, the degree of semantic interference, was signifi-cantly greater in the HL group than in the nor-mal comparison groups in the analysis by subjects, with a significant type x group inter-action : HL-age : F, = 13.74, df = 1, 88, p = .0004 ; HL-vocabulary : F, = 3 .87, df = 1, 88, p = .05 ; HL-phonology : F, = 4.29, df = 1, 58, p = .04 . The type x group interaction did not achieve sig-nificance, however, in the analysis by items .

The pattern of semantic interference across SOA also differed between the HL and normal comparison groups, with a marginally significant interaction between type, SOA, and group : HL-age : F, = 2.80, df = 2, 176, p = .06; F2 = 2.81, df = 2, 28, p = .07 ; HL-vocabulary: F, = 2.54, df = 2, 176, p = .08 ; F2 = 2 .53, df = 2, 28, p = .09 ; HL-phonology : F, = 2.63, df = 2, 116, p = .07 ; F2 = 2.36, df = 2, 28, p = .11 . With regard to the HL group, the difference in related versus unre-lated naming latencies exceeded Fisher's least significant difference (LSD) test (Keppel, 1973) at all SOAs. In the normal comparison groups, on the other hand, the difference in related ver-sus unrelated naming latencies consistently exceeded Fisher's LSD at -150 msec SOA but never achieved the LSD at +150 msec SOA. At 0 msec SOA, the difference in naming times exceeded Fisher's LSD only in the age compari-son group. Previous results in adults also show some variability in the degree of interference at 0 msec SOA (e.g ., compare the results of Schriefers et al, 1990, and Damian and Martin, 1999).

Individual Variability in the HL Group

To probe individual variability in the degree of interference at -150 msec and +150 msec SOAs in the HL group, we conducted multiple regression analyses . The criterion variable was the degree of interference ; the predictor variables were the baseline naming speed, the average degree of HL, the age at identification/amplifi-cation of the loss, and a composite age-related variable . The intercorrelations among this set of variables were minimal, with r2 values ranging

from .01 to .10 (all p values > .05) . The age-related variable was formed by a linear combi-nation of age, receptive vocabulary, expressive vocabulary, articulation proficiency, visual per-ception, and word recognition . The intercorre-lations among the age-related variables were significant, with r2 values ranging from .32 to .57 (all p values <_ .001).

Results showed that the degree of interfer-ence at -150 msec SOA could not be predicted from knowledge of the demographic variables, R2 = .14, omnibus F = 1.04, p = .41. The degree of interference at +150 msec SOA, however, was significantly associated with the set of predictor variables, R2 = .40, omnibus F = 4.21, p = .01. Both the baseline naming speed, partial F = 4.71, p = .04, and the age at identifica-tion/amplification of the loss, partial F = 6.10, p = .02, uniquely accounted for a significant proportion of the variation in interference at +150 msec SOA.

Figure 4 details the degree of interference at each SOA as a function of the baseline naming speed and the age at identification/amplification of the loss . At -150 msec SOA, all subgroups showed interference, as expected from the sta-tistical analysis . At +150 msec SOA, however, semantic interference occurred primarily in the subgroups with an unusually slow baseline nam-ing speed and an early age of identification/ampli-fication. It is also the case that the degree of interference in these subgroups appeared more pronounced at +150 msec SOAthan at-150 msec SOA. The suspected onsets of the loss were roughly comparable in the subgroups, so it is dif-ficult to interpret the differences in the age of iden-tification/amplification . Numerous possible explanations exist, such as progressive hearing losses in the late identification/amplification sub-group . The demographic data indicated that the children who were identified/amplified earlier had slightly more hearing loss, slower baseline naming speeds, and poorer receptive vocabular-ies than the children who were identified/ampli-fied later (about 56 dB HL vs 37 dB HL; 1045 msec vs 825 msec; 18th percentile vs 41st percentile) . With regard to the baseline naming speed sub-groups, the age of identification/amplification in the slowest subgroup was about 1 year earlier than the age in the fastest subgroup . Further studies are necessary to disentangle the factors associated with individual variability in a sensi-tive manner.

To recap, semantic interference in the TD groups occurred when the onset of the distractor was before the onset of the picture (-150 msec)

473


Hearing Impairment

300

200

100

0

Figure 4 Degree of interference at-150 msec and +150 msec stimulus onset asynchronies (SOAs) in the hearing loss group as a function of the baseline naming speed and the age at iden-tification/amplification of the loss . The middle subgroup of each column contains the children with standard scores between ± 0.75; the outermost subgroups contain the children with standard scores of < 0.75 or > 0.75, respectively. The average naming speeds for the subgroups ranged from 627 msec to 1737 msec ; the average ages at identification/amplification ranged from 25 to 69 months .

300

200

100

0

627 883 1737 25 48 69

Baseline Naming Speed (msec) Age ID &Amplification (mo)

but not when the onset of the distractor was after the onset of the picture (+150 msec), mirroring pre-vious findings in normal adults (Schriefers et al, 1990 ; Damian and Martin, 1999). In contrast, semantic interference in the HL group occurred at all SOAs . As seen in Figure 3, the degree of interference (about 100 to 125 msec) was similar in all groups at -150 msec SOA. At +150 msec SOA, however, the pattern of results for the HL and TD groups diverged; naming latencies for the related and unrelated distractors were similar in the TD groups but differed prominently in the HL group, again by about 125 msec . Analysis of indi-vidual variability in the HL group revealed that the children who showed the unexpected inter-ference at +150 msec SOA had unusually slow baseline naming speeds and an early age of iden-tification/amplification .

DISCUSSION

T his research examined the influence of audi-tory semantically related distractors on pic-

ture naming by children with HL with a new children's cross-modal picture-word test . Results showed significant semantic interference in all groups . Similar to adults, accessing the seman-tic content of words was mandatory during lin-guistic processing by young children, even those with HL. This fording agrees with our previous observation of Stroop interference in TD children and children with mild-to-moderate HL (Jerger et al, 1997). The presence of interference effects also indicates that both the TD children and the children with HL appreciated the conceptual categorical relations between the picture-word

pairs. The organization of semantic memory appears highly structured in terms of categori-cal knowledge in these children . Although the items used are early learned and highly famil-iar, the results suggest nonetheless that seman-tic representations are strikingly similar in TD children and children with, on average, moder-ate HL. Thus, early lexical learning appears to remain robust across a range of auditory input experiences.

With regard to the time course of semantic interference, the degree of interference was sim-ilar in the HL and TD groups at -150 msec SOA. At +150 msec SOA, however, the HL group continued to show pronounced interference, whereas the TD groups showed none . In the TD groups, the semantic distractors had an effect only at an early stage of lexical access for pic-ture naming, yielding a time course similar to that seen previously in adults (Schriefers et al, 1990; Damian and Martin, 1999). Apparently, for TD children, as well as adults, there is a point at which a lexical item is selected, and beyond that point, the semantic information of the dis-tractor no longer influences the picture-naming process .

In the presence of childhood HL, however, the time course of semantic interference was abnormally prolonged. An analysis of individual variability in the HL group indicated that all chil-dren showed semantic interference at -150 msec SOA but not at +150 msec SOA. Semantic inter-ference at the latter SOA occurred primarily in the children with HL who had unusually slow baseline naming speeds and an early age of identification/amplification of the loss . With regard to inferences about the time course of

474


semantic activation, the broadly spread pattern of interference implies that the lexical seman-tic stage of processing was abnormally prolonged (see Fig. 1) .

Traditionally, slower processing speeds have been assumed to reflect limitations in process-ing capacity and resources, including less rapid access to object names (Kail, 1996) . Slowed picture-naming times also characterize children with poor reading comprehension in spite of age-appropriate reading accuracy and print decoding skills (Nation et al, 2001) . If we con-sider Gilbertson and Kamhi's (1995) argument that children with mild-to-moderate HL repre-sent a heterogeneous group with respect to lan-guage ability, perhaps it is the HL children with poorer semantic skills who are exhibiting the abnormal time course of semantic interference . The presence of poorer language skills might also account for the slowed baseline naming speed

(slowed lexical access) and the earlier age of identification/amplification . Perhaps the chil-dren who were identified/amplified earlier were the ones exhibiting greater difficulties in real-life communication situations . Although the composite age-related variable did not make a unique contribution to predicting interference at +150 msec SOA, receptive vocabulary ability was noticeably poorer in the subgroup with an earlier age of identification/amplification and interference versus the subgroup with a later age of identification/amplification and no interference (see Fig . 4), as noted previously. The range of per-centile scores in the subgroups overlapped, how-

ever, so we must again caution that further data are necessary to unravel the factors producing a broader window of lexical semantic activation.

It is also important to note that absolute slow-ing per se cannot account for the abnormal time course of semantic interference in the HL group. Results in Figures 2 and 3 clearly show that the TD groups had pronounced variations in absolute naming times that did not alter the time course of semantic interference . In fact, average absolute baseline naming times in the input phonology group were not dramatically different from those in the slowest subgroup of older children with HL and interference at +150 msec SOA (see Fig. 4). This observation is consistent with Lahey and Edwards' (1996) important idea that we should consider the speed of each individual stage of pro-cessing in addition to overall speed per se . Our results imply that the lexical semantic stage of pro-cessing contributed to the slower naming times in the children with HL but not in the normal younger children in the phonology group.

Two other explanations for the abnormally prolonged lexical semantic stage of processing should also be considered . One concerns a pos-sible delay in development of the capacity for inhibition and resistance to interference (Jerger et al, 1997, 1999) . Inhibition refers to the abil-

ity to inhibit a response, whereas resistance to interference refers to the ability to ignore irrel-evant information (Bjorklund, 1995) . Delayed inhibitory skills and resistance to interference would render the distractor more difficult to ignore, perhaps increasing competition between coactivated picture-word lexical representations and hence slowing lexical access . Immaturity of these skills is also consistent with the trend toward a greater degree of semantic interference in the HL group (see Figs . 3 and 4) . Thus, an explanation based on delayed development of inhibition and resistance to interference seems consistent with the overall evidence .

The other possible explanation concerns the evidence that children with auditory dis-crimination difficulties may rely dispropor-tionately on semantic processes (Rodda and Grove, 1987 ; Stackhouse and Wells, 1997) . For example, false recognition errors on tests requir-ing new/old judgments to each item of a to-be-remembered list indicate the use of a semantic code in children with HL (falsely say "yes" to "pony" when "horse" was the previous item) and a phonologic code in normal children (falsely say "yes" to "boat" when "goat" was the previ-ous item) (Frumkin and Anisfeld, 1977). Possi-ble reasons that children with HL may rely disproportionately on semantic coding for pro-cessing is that (1) the formation of temporary phonologic representations for learning and remembering (Gathercole and Baddeley, 1993) is less efficient (Gilbertson and Kamhi, 1995 ; Briscoe et al, 2001) and (2) syntactic knowl-edge is relatively impoverished (Swisher, 1976 ; Yoshinaga-Itano, 1997). Both phenomena might promote prolonged semantic activation by increased reliance on word meaning and world knowledge for remembering and comprehend-ing. Overall, however, an explanation based on increased reliance on semantic processes does not seem consistent with all of the evidence . It is not immediately apparent, for example, why a normal, but abnormally prolonged, lexical semantic stage of activation would produce abnormally slowed baseline naming speeds . Slowed baseline naming times seem more con-sistent with slowed access to object names owing to limitations in processing capacities and skills . It will be important for further studies to assess

475


and differentiate the effects on picture-word results of differences in semantic skills, phono-logic skills, capacity for inhibition, and resis-tance to interference .

Acknowledgment. This work was supported by the National Institute on Deafness and Other Communication Disorders, grant DC-00421, to the University of Texas at Dallas . Part of the work was carried out while the first author was at Baylor College of Medicine . We are grate-ful for the assistance of Aron Barbey, Janette Cross, Tara Gaspar, Marlaina Martin, Ingrid McCloud, Eric Wood (data collection), Justin Rachels (computer program-ming), and Letitia Pelzer, MD (consultant on oral-motor function). We appreciate the collegial support of the Callier Center for Communication Disorders, particularly Dr. Ross Roeser and Ms . Karen Clark. Finally, we give a spe-cial thank you to two anonymous reviewers for their consultative and constructive comments .

Davis J, Elfenbein J, Schum R, Bender R. (1986) . Effects of mild and moderate hearing impairments on language, educational, and psychosocial behavior of children . J Speech Hear Disord 51:53-62 .

Dell G. (1986) . A spreading activation theory of retrieval in language production . Psychol Rev 93:283-321 .

Dell G, O'Seaghdha P. (1991) . Mediated and convergent lexical priming in language production : a comment on Levelt et al . (1991) . Psychol Rev 98:604-614 .

Dunn L, Dunn L. (1997) . The Peabody Picture Vocabulary Test . 3rd Ed . Circle Pines, MN: American Guidance Service.

Francis W, Kucera H. (1982). Frequency Analysis ofEnglish Usage: Lexicon and Grammar. Boston : Houghton Mifflin.

Frankenburg W Doorninck W von, Liddell T, Dick N. (1976) . Denver Prescreening Developmental Questionnaire. Denver, CO: Ladoca Publishing Foundation.

REFERENCES

American National Standards Institute . (1989) . American Standard Specifications for Audiometers. (ANSI 83.6-1989). New York : ANSI.

Ayres J. (1978) . Southern California Figure Ground Visual Perception Test . 6th Ed. Los Angeles : Western Psychological Services .

Beery K. (1989) . The Developmental Test of Visual Motor Integration. 3rd Ed. Cleveland: Modern Curriculum Press.

Bjorklund D . (1995). Children's Thinking. Pacific Grove, CA : Brooks/Cole .

Blair J, Petersen M, Viehweg S. (1985) . The effects of mild sensorineural hearing loss on academic performance of young school-age children . Volta Rev 87:87-93 .

Briscoe J, Bishop D, Norbury C . (2001) . Phonological pro-cessing, language, and literacy: a comparison of children with mild-to-moderate sensorineural hearing loss and those with specific language impairment. J Child Psychol Psychiatry 42:329-340 .

Campbell R, Wright H. (1990) . Deafness and immediate memory for pictures : dissociations between "inner speech" and the "inner ear"? J Exp Child Psychol 50:259-286 .

Carroll J, White M. (1973a). Age-of-acquisition norms for 220 picturable nouns. J Verbal Learn Verbal Behav 12:563-576 .

Carroll J, White M. (1973b). Word frequency and age of acquisition as determiners of picture-naming latency. Q J Exp Psychol 25 :85-95 .

Cutting J, Ferreira V (1999) . Semantic and phonologi-cal information flow in the production lexicon. J Exp Psychol Learn Mem Cogn 25 :318-344 .

Damian M, Martin R. (1999) . Semantic and phonologi-cal codes interact in single word production . J Exp Psychol Learn Mem Cogn 25:345-361 .

Damian M, Vigliocco G, Levelt W (2001) . Effects of seman-tic context in the naming of pictures and words. Cognition 81:77-86 .

Frumkin B, Anisfeld M. (1977) . Semantic and surface codes in the memory of deaf children . Cogn Psychol 9:475-493 .

Gathercole S, BaddeleyA. (1993) . Working Memory and Language. Hillsdale, NJ : Lawrence Erlbaum.

Gilbertson M, Kamhi A. (1995) . Novel word learning in children with hearing impairment . J Speech Hear Res 38:630-642 .

Gilhooly K, Logie R. (1980) . Age-of-acquisition, imagery, concreteness, familiarity, and amgibuity measures for 1944 words. Behau Res Methods Instr 12:395-427 .

Glaser W. (1992) . Picture naming . Cognition 42:61-105 .

Goldman R, Fristoe M. (1986) . Goldman Fristoe Test of Articulation (Revised). Circle Pines, MN: American Guidance Service.

Goodenough F. (1935) . The development of the reactive process from early childhood to maturity. J Exp Psychol 18:431-450 .

Hoemann H, Andrews C, DeRosa D. (1974). Categorical encoding in short-term memory by deaf and hearing chil-dren . J Speech Hear Res 17:426-431 .

Hollingshead A. (1975) . Four Factor Index of Social Status. New Haven, CT: Department of Sociology, Yale University.

Jerger S, Martin R, Pirozzolo F. (1988) . A developmen-tal study of the auditory Stroop effect . Brain Lang 35:86-104.

Jerger S, Grimes A, Tran T, et al . (1997) . Childhood hearing impairment : processing dependencies in multi-dimensional speech perception for an auditory level of analysis . Ear Hear 18:513-535 .

Jerger S, Pearson D, Spence M. (1999) . Development-al course of auditory processing interactions : Garner interference and Simon interference . J Exp Child Psychol 74:44-67 .

Jerger S, Lai L, Marchman VA . (2002a). Picture naming by children with hearing loss : II . Effect of phonologically related auditory distractors. J Am Acad Audiol 13 :478-492 .

476

Jerger S, Martin R, Damian M. (2002b). Semantic and phonological influences on picture naming by children and teenagers. J Mem Lang 47:229-249 .

Jescheniak J, Schriefers H. (1998) . Discrete serial versus cascaded processing in lexical access in speech produc-tion : further evidence from the coactivation of near-synonyms. J Exp Psychol Learn Mem Cogn 24 : 1256-1274.

Kail R. (1996) . Nature and consequences of developmental change in speed of processing . Swiss J Psychol 55:133-138 .

Keppel G . (1973) . Design & Analysis . A Researcher's Handbook . 2nd Ed . Englewood Cliffs, NJ : Prentice-Hall .

La Heij W Dirkx J, Kramer P. (1990) . Categorical inter-ference and associative priming in picture naming. Br J Psychol 81:511-525 .

Lahey M, Edwards J. (1996) . Why do children with spe-cific language impairment name pictures more slowly than their peers? J Speech Hear Res 39:1081-1098 .

LeCompte D, Neely C, Wilson J. (1997). Irrelevant speech and irrelevant tones: the relative importance of speech to the irrelevant speech effect . JExp Psychol Learn Mem Cogn 23:472-483 .

Levelt W Schriefers H, Vorberg D, et al. (1991) . The time course of lexical access in speech production : a study of picture naming. Psychol Rev 98:122-142 .

Levelt W, Roelofs A, Meyer A. (1999) . A theory of lexical access in speech production . Behau Brain Sci 22:1-75.

MacWhinney B . (1993) . The CHILDES Database . 2nd Ed . Dublin, OH : Discovery Systems .

Martin R, Wogalter M, Forlano J. (1988) . Reading com-prehension in the presence of unattended speech and music. J Mem Lang 27:382-398.

Meyer A, Schriefers H. (1991) . Phonological facilitation in picture-word interference experiments: effects of stim-ulus onset asynchrony and types of interfering stimuli. J Exp Psychol Learn Mem Cogn 17:1146-1160 .

Nation K, Marshall C, Snowling M. (2001) . Phonological and semantic contributions to children's picture naming : evidence from children with developmental reading dis-orders . Lang Cogn Processes 16:241-259 .

Nusbaum H, Pisoni D, Davis C. (1984) . Sizing Up the Hoosier Mental Lexicon: Measuring the Familiarity of 20,000 Words. Research on Speech Perception: Progress Report no. 10 . Bloomington, IN : Indiana University.

Nusbaum H, Pisoni D . (1985) . Constraints on the per-ception of synthetic speech generated by rule . Behav Res Methods Instr Comp 17:235-242 .

Osberger M, Hesketh L. (1988) . Speech and language disorders related to hearing impairment . In : Lass N, Northern J, McReynolds L, Yoder D, eds. Handbook of Speech-Language Pathology and Audiology. Toronto: BC Decker, 858-886.

Paivio A, Yuille J, Madigan S . (1968) . Concreteness, imagery, and meaningfulness values for 925 nouns . J Exp Psychol 76(Suppl):1-25.


Pedhazur E. (1982) . Multiple Regression in Behavioral Research. Explanation and Prediction . 2nd Ed . New York : Holt, Rinehart & Winston.

Peltzer L. (1997) . Oral Motor Function Questionnaire. Houston: Baylor College of Medicine .

Rader K. (1977) . Rader Near Point Vision Test . Tulsa, OK : Modern Education Corp .

Rayner K, Springer C. (1986) . Graphemic and semantic similarity effects in the picture-word interference task . Br J Psychol 77 :207-222 .

Rodda M, Grove C. (1987) . Language, Cognition and Deafness . Hillsdale, NJ : Lawrence Erlbaum.

Roelofs A. (1992) . A spreading-activation theory of lemma retrieval in speaking . Cognition 42:107-142 .

Ross M, Lerman J. (1971) . Word Intelligibility by Picture Identification . Pittsburgh : Stanwix House.

Ruffin-Simon C. (1983) . Relationship between produc-tion and perception of speech : some developmental issues . In : Hochberg I, Levitt H, Osberger M, eds. Speech of the Hearing Impaired. Research, Training, and Personnel Preparation. Baltimore: University Park Press, 233-249.

Schriefers H, Meyer A, Levelt W. (1990) . Exploring the time course of lexical access in language production : pic-ture-word interference studies. J Mem Lang 29:86-102.

Snodgrass J, Vanderwart M. (1980) . A standardized set of 260 pictures : norms for name agreement, image agree-ment, familiarity, and visual complexity. J Exp Psychol Hum Learn Mem 8:174-215 .

Staekhouse J, Wells B. (1997) . Children's Speech and Literacy Difficulties . APsycholinguistic Framework. San Diego: Singular.

Starreveld P, LaHeij W. (1996) . Time-course analysis of semantic and orthographic context effects in picture naming . J Exp Psychol Learn Mem Cogn 22:896-918 .

Swisher L. (1976) . The language performance of the oral deaf. In : Whitaker H, Whitaker H, eds. Studies in Neurolinguistics . Vol. 2 . New York : Academic Press, 59-93.

'rye-Murray N. (1992) . Articulatory organizational strate-gies and the roles of audition . Volta Rev 94:243-260 .

Tye-Murray N, Spencer L, Gilbert-Bedia E . (1995) . Relationships between speech production and speech per-ception skills in young cochlear-implant users. JAcoust Soc Am 98:2454-2460 .

Wechsler D. (1974) . Wechsler Intelligence Scale for Children-Revised . New York : The Psychological Corporation.

Williams K. (1997) . Expressive Vocabulary Test . Circle Pines, MN : American Guidance Service.

Yoshinaga-Itano C. (1997) . The challenge of assessing language in children with hearing loss . Lang Speech Hear Serv Schools 28:362-373 .

picture naming by children with hearing loss: i. effect … · j am acad audiol 13: 463-477 (2002)...

Documents