Concept vocabulary in children who are deaf or hard of hearing

Abstract

This longitudinal study compares concept vocabulary knowledge of children with cochlear implants (n = 40) and children with hearing aids (n = 30) to that of their typical hearing peers (n = 40). Participants completed the Bracken Basic Concept Scale: Expressive (BBCS:E) at ages 4 and 6. Results revealed significant differences in concept vocabulary knowledge between both groups of children who are deaf or hard of hearing (DHH) and the typical hearing group. Although all groups improved BBCS:E test performance between ages 4 and 6, the rate of improvement in children who are DHH did not trend toward catching up over time. Omnibus expressive vocabulary outcomes predicted BBCS:E performance, but age of amplification did not. These preliminary data suggest persistence in concept vocabulary deficits in children who are DHH and developing spoken language, at least through entry into elementary school.

Introduction

Children who are deaf and hard of hearing (DHH), on average, tend to know fewer words than their peers with typical hearing, possibly related to barriers to language exposure (Berger et al., 2024; Convertino et al., 2014) This pattern is consistent for children who are learning spoken language through a hearing aid or cochlear implant: They tend to know fewer words than their peers with typical hearing (e.g., Geers et al., 2009; Lund, 2016). However, some extant literature also indicates that children who are DHH may know different words than children with typical hearing as well (e.g., Bracken & Cato, 1986; Lund, 2019; Rush et al., 2023) Within the vocabulary literature, there have been a few studies that suggest that children who are DHH may present with persistent deficits in concept vocabulary knowledge in comparison to their same-age, typical hearing peers (Bracken & Cato, 1986; Brenza et al., 1981; Davis, 1974). In this study, we use the term concept vocabulary to refer to functional words that represent abstract yet basic concepts, such as size/comparison terms, direction/position terms, and quantity terms. Because concept vocabulary knowledge highly correlates with measures of academic success (e.g., Howell & Bracken, 1992; Panter, 2000), this potentially persistent gap in concept vocabulary knowledge could have substantial academic consequences for children who are DHH and learning spoken language. All existing studies of concept knowledge in children who are DHH compare performance on concept vocabulary measures between different age groups. None of these studies track performance on this skill over time within the same group of children. Additionally, all existing studies were conducted before or during the 1980s (Bracken & Cato, 1986; Brenza et al., 1981; Davis, 1974), prior to substantial advances in hearing technology from which children who are DHH today benefit. These technological advances could theoretically impact concept word knowledge (e.g., Davis, 1974). Thus, the purpose of this longitudinal study is to evaluate concept vocabulary knowledge over time in children who are DHH who use current hearing technology.

Concept vocabulary knowledge

Concept words, which are typically categorized into broader concepts or categories, include colors, letters, numbers, size/comparisons (e.g., big, small, tall), shapes (e.g., curve, angle, triangle, cube), direction/position (e.g., under, outside, far), self-/social-awareness (e.g., happy, sick, mother, young), texture (e.g., solid, smooth, dry), quantity (e.g., piece, full, most), and time/sequence words (e.g., twice, first, late; Bracken & Crawford, 2010). Children begin using some of the most basic concept words (e.g., up, more) as early as 18 months. As their language matures, children use concept words in phrases and sentences to communicate messages of increasing complexity (e.g., mommy up, the book is on the table; my cup is empty, but yours is full). By six years of age, children have learned most concept words but still acquire a few later-developing words well into their school-age years (for review on development, see Boehm, 1991).

Concept vocabulary knowledge is essential to academic success. Early childhood education standards across all 50 states specify that children must learn concept words (Bracken & Crawford, 2010). Children use their knowledge of concept vocabulary to learn and follow classroom instructions. For example, greater than, less than, and equal to are size/comparison concept words and half and quarter are quantity concept words (Bracken & Crawford, 2010), all of which are used in mathematical instruction. The concept words beginning and end are direction/position terms (Bracken & Crawford, 2010) that are relevant to reading and writing stories. Students must understand the concept words down and under to follow the instructions “put your pencil down” and “hang your backpack under your name.” These few examples illustrate the central role that concept words play in everyday academic instruction.

As one might expect then, concept vocabulary knowledge is correlated with other omnibus outcomes measures like vocabulary development (Breen, 1985) and performance on tests of intelligence (Howell & Bracken, 1992) and achievement (Breen, 1985; Piersel & McAndrews, 1982; Zucker & Riordan, 1990). Additionally, concept vocabulary knowledge has been correlated with tests of readiness to learn (Panter, 2000; Piersel & McAndrews, 1982), specifically readiness for instruction in math and reading (Panter, 2000). Furthermore, concept words are used in directions and questions on various academic tests, including tests of intelligence (Bracken, 1986; Flanagan et al., 1995) and achievement (Cummings & Nelson, 1980). Thus, knowledge of these words likely impacts ability to follow directions on these tests, which could impact performance and test results. In conclusion, concept vocabulary knowledge plays an important role in learning and performance in the academic setting.

Vocabulary in Children Who Are Deaf or Hard of Hearing

It is known that children who are DHH, on average, present with gaps in expressive and receptive vocabulary knowledge in comparison to their typical hearing, same-age peers (for review, see Lund, 2016). Most studies that compare vocabulary knowledge of children who are DHH to that of their typical hearing peers use standardized measures (Lund, 2016). These measures evaluate breadth of vocabulary knowledge (Brownell, 2011; Dunn & Dunn, 2007), but they provide little information about types of words known (e.g., concept words).

Three studies have compared concept vocabulary knowledge of children who are DHH to that of their typical hearing peers (Bracken & Cato, 1986; Brenza et al., 1981; Davis, 1974). Together, these studies suggest that children who are DHH do know some concept words. However, they know fewer words than their typical hearing, age-matched peers (e.g., Bracken & Cato, 1986; Davis, 1974), and this gap may persist over time (Brenza et al., 1981).

Bracken and Cato (1986) used the Bracken Basic Concept Scale (Bracken, 1984) to compare concept knowledge of children who are DHH (n = 17) to that of their age-matched peers (n = 17). About half of the children who are DHH communicated using oral communication, and the other half used total communication. The Bracken Basic Concept Scale is a standardized test that measures knowledge of 258 concept words across seven categories: School Readiness Composite, Direction/Position, Social/Emotional, Size, Texture/Material, Quantity, and Time/Sequence. Results revealed that children who are DHH had significantly lower test scores than the typical hearing group across all categories (all p < .001). Children who are DHH performed, on average, 2 SD below their typical hearing, age-matched peers: a substantial difference.

Davis (1974) compared concept knowledge of 24 children who are DHH to that of their age-matched peers (n = 24). The type of communication (e.g., oral, sign) used by the children who are DHH was not reported. Participants from both hearing status groups were between the ages of 6;0 and 8;11. Evaluators administered the Boehm Test of Basic Concepts, which evaluates knowledge of 50 concept vocabulary words from four categories: space, time, quantity, and miscellaneous (e.g., different, matches). Authors compared results across age groups and audibility (the latter comparison within the DHH group only).

Within the DHH and typical hearing groups, children were further grouped by age: 6-year-olds, 7-year-olds, and 8-year-olds. Although raw test scores for older age groups were slightly higher than those for younger age groups in both hearing status groups, significant differences between age groups were not found, suggesting limited growth past age 6. Of particular concern, however, is that (although it was not statistically compared) the typical hearing group raw scores were substantially higher than the DHH group raw scores for each age group: 17-point difference for 6-year-olds, 16-point difference for 7-year-olds, and 13.33-point difference for 8-year-olds. Together, these findings suggest that children who are DHH are behind their typical peers in concept vocabulary development past age 6 (Davis, 1974), which is the age that children are expected to have learned most concept words (Boehm, 1991).

Furthermore, Davis (1974) evaluated the impact of audibility on concept knowledge test performance. The DHH group was further split into two groups: pure tone average between 35 and 50 dB and pure tone average between 51 and 70 dB (n = 12 for both groups). Children with greater pure tone averages performed significantly worse than children with lower pure tone averages (p < .0001), suggesting that access to sound may be an influential factor in concept word knowledge for children who are DHH.

Finally, Brenza et al. (1981) evaluated the performance of 13- to 14-year-old adolescents (n = 15) who are profoundly deaf on the Boehm Test of Basic Concepts. All participants were oral communicators. Consistent with other studies, participants’ test results were significantly lower than those of typical-hearing children from the test’s normative data, revealing deficits in concept knowledge for this population. Of particular interest to the current study, authors compared these participants’ results to the results of the 6- to 8-year-old participants from Davis’s (1974) study and found no significant differences in performance, suggesting limited growth in concept word knowledge between these age groups. Together, these findings suggest that children who are DHH are behind their typical peers in concept vocabulary development (Bracken & Cato, 1986; Brenza et al., 1981; Davis, 1974), and they may not catch up over time (Brenza et al., 1981).

These three studies reveal deficits in concept vocabulary knowledge in children who are DHH (Bracken & Cato, 1986; Brenza et al., 1981; Davis, 1974), but the studies all present a common limitation: They were conducted in the mid-1970s and 1980s. Hearing technology has advanced in multiple ways since the 1980s (e.g., use of cochlear implants in children, ASHA, 2020; better quality of amplification in hearing aids, U.S. Department of Health and Human Services, 2022). These advancements provide children who are DHH with better access to sound (ASHA, n.d.; U.S. Department of Health and Human Services, 2022) and better access to sound is associated with better vocabulary outcomes in children who are DHH (e.g., Davidson et al., 2014; Davis, 1974; Tomblin et al., 2015). Thus, these three studies may not be representative of current concept vocabulary outcomes in children who are DHH. In light of advances in hearing technology, we must reevaluate this populations’ concept vocabulary knowledge to determine whether they still present with deficits in concept vocabulary knowledge.

Age at amplification

Technological improvements have also contributed to our understanding of the role of age at amplification in spoken language outcomes. Children with early amplification fittings, which include hearing aids for children with less severe losses and cochlear implants for children with more severe degrees of hearing loss, tend to have higher omnibus language outcomes than children with later fittings (e.g., Ching et al., 2017; Tomblin et al., 2015). Ching et al. (2017) conducted a prospective study of children with a range of degrees of hearing loss (N = 350) to elucidate the role of amplification fitting. For children with degrees of hearing loss that could be addressed by a hearing aid, children who received hearing aids early (for example, at 3 months) had higher language performance than those who received amplification later (for example, at 24 months). Similarly, children whose hearing loss was addressed by a cochlear implant had better outcomes at age 5 if they received a cochlear implant prior to age 1 as compared to 24 months. Findings like these have driven Early Hearing Detection and Intervention program guidelines, which advise that babies should receive a hearing screening by 1 month, a hearing loss diagnosis by 3 months, and enroll in early intervention by 6 months (JCIH, 2019). However, studies to date do not indicate the influence of an audiological variable like age at amplification fitting on specific vocabulary outcomes. This information would be important for determining whether children who are fitted sufficiently early are likely to bypass difficulties seen in other children who are DHH.

Current study

In conclusion, few studies have evaluated concept knowledge in children who are DHH in comparison to their typical-hearing, same-age peers. All results thus far reveal deficits in concept word knowledge in children who are DHH (Bracken & Cato, 1986; Brenza et al., 1981). Additionally, existing (albeit limited) data suggest that concept vocabulary development may plateau at about age 6 for both typical hearing children (e.g., Boehm, 1991; Davis, 1974) and children who are DHH (Brenza et al., 1981; Davis, 1974), potentially yielding persistent deficits for children who are DHH (Brenza et al., 1981) if educators assume that by age 6, concept vocabulary no longer needs to be taught. However, none of these studies include a longitudinal design in which the same children are tracked over time to describe concept vocabulary development. Also, none of these studies measure concept vocabulary knowledge in children who use current hearing technology. Furthermore, other important audiological factors, like age of amplification fitting, have yet to be considered as a predictor in concept vocabulary outcomes for children today.

We need to understand how concept vocabulary knowledge develops in children who are DHH who use current hearing technology. Such data can inform future investigations that evaluate methods of closing this concept word knowledge gap. Thus, the purpose of this longitudinal study is to evaluate concept vocabulary knowledge over time in children who are DHH. The research questions include the following:

1. Given technological improvements in children over the past decades, do children with cochlear implants and hearing aids continue to perform lower on a single-word measure of conceptual vocabulary as compared to children with typical hearing?

2. Do children with cochlear implants and children with hearing aids improve their concept vocabulary knowledge at the same rate as children with typical hearing between the ages of four and six?

3. Does age at amplification fitting and omnibus vocabulary knowledge predict concept vocabulary knowledge outcomes?

Method

This study was approved by the University of South Carolina Internal Review Board (IRB of record with Texas Christian University in agreement). Participants’ parents/guardians provided informed consent, and participants provided assent prior to completing study procedures. Data reported in this manuscript come from a larger longitudinal study (National Institute on Deafness and Other Communication Disorders [NIDCD]/National Institutes of Health [NIH] R01; PI: Werfel & Lund) that investigates early language and literacy acquisition in children who are DHH and their typical hearing peers.

Participants

A total of 110 children participated in this study, all of whom fell into one of three groups: children who are DHH who use cochlear implants (CI; n = 40), children who are DHH who use hearing aids (HA; n = 30), and a typical hearing group matched for age (TH; n = 40). Five participants wore both a cochlear implant and a hearing aid. These children were grouped into the cochlear implant group.

All participants communicated primarily using spoken language (i.e., did not use primarily sign language to communicate, although some children were exposed to sign support, such as sign-supported English, to varying degrees at points throughout their development). Participants spoke primarily English in the home and were from multiple states in the continental United States, spanning the Northeast, Midwest, South, and West regions. See Table 1 for additional descriptive data about participants.

Measures

Descriptive measures

Several measures were used to characterize the overall skills of children in this study. The Expressive One-Word Picture Vocabulary Test–Fourth Edition (EOWPVT-4; Brownell, 2011) was used to measure expressive vocabulary. The test booklet displayed single images, and the evaluator asked participants to name each image. Raw scores were converted to standardized scores according to test manual procedures.

The Peabody Picture Vocabulary Test–Fourth Edition (PPVT; Dunn & Dunn, 2007) was used to measure receptive vocabulary. The test booklet displayed an array of four images, and the evaluator asked participants to point to the named image. Raw scores were converted to standardized scores according to test manual procedures.

The Primary Test of Nonverbal Intelligence (Ehrler & McGhee, 2008) was used to measure nonverbal intelligence. The test booklet displayed an array of line drawings, and the evaluator asked participants to point to the drawing that did not belong with the other drawings. Raw scores were converted to standardized scores according to test manual procedures.

The Test of Early Language Development–Third Edition (TELD-3; Hresko et al., 1999) was used to measure expressive and receptive language. Evaluators followed administration procedures for both the expressive and receptive portions of the test. Raw scores were converted to standardized scores according to test manual procedures.

Concept vocabulary measure

The Bracken Basic Concept Scale: Expressive (BBCS:E; Bracken, 2006) was used to measure expressive concept vocabulary knowledge. The test includes 10 subtests: Colors, Letters/Sounds, Numbers/Counting, Sizes/Comparisons, Shapes, Direction/Position, Self-/Social Awareness, Texture/Material, Quantity, and Time/Sequence. The Colors subtest evaluates knowledge of primary and basic colors. The Letters/Sounds subtest evaluates knowledge of uppercase and lowercase letter names and corresponding letter sounds. The Numbers/Counting subtest evaluates knowledge of single-digit and double-digit numbers and ability to count the number of objects in a group. The Sizes/Comparisons subtest evaluates knowledge of concept words that describe things in one, two, or three dimensions (e.g., tall, short, and big, respectively). Additionally, it measures the ability to compare objects based on characteristics. The Shapes subtest evaluates knowledge of two-dimensional shapes (e.g., triangle) and three-dimensional shapes (e.g., pyramid). The Direction/Position subtest evaluates knowledge of concept words that describe location relative to other objects (e.g., under), position relative to self (e.g., open), and direction (e.g., center). The Self-/Social Awareness subtest measures knowledge of concept words related to one’s emotional state (e.g., angry) and words related to kinship, gender, ages relative to another (e.g., old), and social appropriateness (e.g., wrong). The Texture/Material subtest evaluates knowledge of concept words related to object attributes/characteristics (e.g., hot) and an object’s composition (e.g., glass). The Quantity subtest evaluates knowledge of concept words that describe existence, space occupied, and amount (e.g., full). Finally, the Time/Sequence subtest measures knowledge of concept words that describe temporal or sequential occurrences (e.g., first; Bracken, 2006).

The first five subtests (Colors, Letters/Sounds, Numbers/Counting, Sizes/Comparisons, and Shapes) make up the School Readiness Composite. These five tests assess knowledge of concept words that are generally acquired prior to formal education. The remaining five subtests stand alone. When administering subtests that fall within the School Readiness Composite, evaluators begin each subtest at the first question and continue testing until the child meets the discontinue rule (four consecutive wrong responses) or reaches the end of that subtest. After completing subtests, a School Readiness Composite score is calculated, which determines the starting point for the remaining five subtests. Each subtest includes between 7 and 30 questions. The test booklet displays an image that coordinates with a single test item. The examiner reads the prompt for a single test question (e.g., “This rock is small, and this rock is…”; Bracken, 2006, p. 12). The child responds to the test question by verbalizing the concept word that accurately completes the sentence or answers the question. Raw scores are calculated for each subtest and converted into scaled scores using the manual.

Procedures

Participants in the longitudinal study completed a battery of language, literacy, and other academic-related assessments (approximately 20 assessments) every 6 months from ages 4 to 6 and then annually from first to fifth grade. The assessment battery included the descriptive measures and the BBCS:E. Participants completed the BBCS:E at ages 4 and 6.

After all BBCS:E exams were completed, all scores were scored once by a trained research lab member and verified by a second trained research lab member. All data were entered into excel, and over 30% of data entries verified by a second lab assistant. No data entry errors were observed by the second lab assistant, so original entries were used in the following analyses.

Analysis

To address research questions one and two, authors planned six generalized linear mixed models to evaluate the impact of amplification group (TH, HA, and CI) and time (Time 1 completed at age 4 and Time 2 completed at age 6) on BBCS:E subtest scores. The dependent variable for each generalized linear mixed model was the mean score for one of the following BBCS:E subtests: (1) School Readiness Composite (the sum of the raw scores of the Colors, Letters/Sounds, Numbers/Counting, Size/Comparisons, and Shapes subtests), (2) Direction/Position, (3) Self-/Social Awareness, (4) Texture/Material, (5) Quantity, and (6) Time/Sequence. If a participant was missing data for a specific subtest, that participant was excluded from the analysis for that subtest. However, that participant remained in analyses for other subtests if all data were present for those analyses. The main effect of amplification group with follow-up pairwise contrasts using a sequential Bonferroni-adjusted p-value answered the first research question. The main effect of time and an interaction analysis between group and time answered the second research question.

To address research question three, authors planned a multiple linear regression to evaluate the influence of age at amplification (in months) and general expressive vocabulary knowledge on overall BBCS:E test performance. Age at amplification reflected the age at which participants who are DHH received amplification technology (i.e., hearings aids for participants in the HA group and cochlear implants for children in the CI group). Expressive vocabulary knowledge was measured by the EOWPVT-4 raw score that was collected at time 2. The dependent variable, overall BBCS:E test performance, was the BBCS:E Total Composite raw score, meaning that it was the sum of participants’ raw scores from each BBCS:E subtest.

Results

Amplification status (between subjects) and time (within subjects) were included in the model as fixed effects. Participant code was added to the model as a random effect. However, with this random effect in the model, the final Hessian matrix was not positive definite, suggesting that the model may be too complex. Excluding the random effect from the model caused minimal impact on the fit of the model across all analyses (as measured by examination of the Akaike-corrected criterion value). Thus, the random effect was removed from the models, yielding a positive-definite Hessian.

Because the dependent variables are continuous, skew left (as a result of a ceiling effect), and contain some zeros, the best fit for the linear dependent variables was the linear model target distribution. The linear model assumes a normal distribution. Because our dependent variables are slightly skewed, we used robust estimation to handle violations of model assumptions. Follow-up pairwise contrasts were conducted using sequential Bonferroni adjustment to evaluate differences between amplification groups means and time point means.

Research question 1

The first research question asked whether children with cochlear implants and children with hearing aids perform lower than children with typical hearing on measures of concept vocabulary knowledge. See Table 2 for results. On the School Readiness Composite, the full model revealed a significant main effect of amplification group [F(2, 214) = 5.76, p = .004]. Follow-up pairwise contrasts revealed a significant difference in performance between children with TH and CI (Bonferroni-adjusted p = .003) but no significant difference in performance between children with TH and HA (adjusted p = .472) nor between children with HA and CI (adjusted p = .058).

On the Direction/Position subtest, the full model revealed a significant main effect of amplification group [F(2, 225) = 16.68, p < .001]. Follow-up pairwise contrasts revealed significant differences in performance between children with TH and HA (adjusted p = .002) and between children with TH and CI (adjusted p < .001). Children with HA and CI did not perform differently (adjusted p = .157).

On the Self-/Social Awareness subtest, the full model revealed a significant main effect of amplification group [F(2, 214) = 18.40, p < .001]. Follow-up pairwise contrasts revealed significant differences in performance between children with TH and HA (adjusted p = .001) and children with TH and CI (adjusted p < .001) but no difference between children with HA and CI (adjusted p = .159).

On the Texture/Material subtest, the full model revealed a significant main effect of amplification group [F(2, 214) = 16.70, p < .001). Follow-up pairwise contrasts revealed significant differences in performance between children with TH and HA (adjusted p = .006) and children with TH and CI (adjusted p < .001) but no difference between children with HA and CI (adjusted p = .084).

On the Quantity subtest, the full model revealed a significant main effect of amplification group [F(2, 212) = 40.96, p < .001]. Follow-up pairwise contrasts revealed significant differences in performance between children with TH and HA, children with TH and CI (both adjusted p < .001), and children with HA and CI (adjusted p = .006).

Finally, on the Time/Sequence subtest, the full model revealed a significant main effect of amplification group [F(2, 214) = 26.23, p < .001]. Follow-up pairwise contrasts revealed significant differences in performance between children with TH and HA and children with TH and CI (both adjusted p < .001), but no difference between children with HA and CI (adjusted p = .064).

In summary, children with TH outperformed children with CI across all subtests, and children with TH outperformed children with HA on all but the School Readiness Composite, which comprises the earliest developing concept words. Thus, to answer the first research question, children with CI and HA perform lower on single-word measures of conceptual vocabulary as compared to children with TH, consistent with prior work from the 1980s (Bracken & Cato, 1986; Brenza et al., 1981; Davis, 1974).

Research question 2

The second research question asked whether children with cochlear implants and children with hearing aids improve concept vocabulary knowledge at the same rate as children with typical hearing between the ages of 4 and 6. See Table 2 for results. On the School Readiness Composite, the full model revealed a significant main effect of time [F(1, 214) = 177.59, p < .001] but no interaction effect between group and time [F(2, 214) = 0.21, p = .814]. Thus, for the concept vocabulary evaluated in the School Readiness Composite, all groups performed better at age 6 than age 4, but the rate of growth did not vary between groups. Results suggest that children with CI (who perform more poorly than children with TH on this task: adjusted p = .003) do not catch up to their TH peers on early developing concept vocabulary words by age 6.

On the Direction/Position subtest, the full model revealed a significant main effect of time [F(1, 225) = 88.41, p < .001] but no interaction effect between group and time [F(2, 225) = 1.76, p = .174]. On the Self-/Social Awareness subtest, the full model revealed a significant main effect of time [F(1, 214) = 120.96, p < .001] but no interaction effect between group and time [F(2, 214) = 1.62, p = .200]. On the Texture/Material subtest, the full model revealed a significant main effect of time [F(1, 214) = 118.54, p < .001] but no interaction effect between group and time [F(2, 214) = .03, p = .971]. On the Time/Sequence subtest, the full model revealed a significant main effect of time [F(1, 214) = 146.68, p < .001] but no interaction effect between group and time [F(2, 214) = 0.133, p = .893].

For each of those four models (for subtests Direction/Position, Self-/Social Awareness, Texture/Material, and Time/Sequence), children with TH performed significantly better than children with HA and children with CI (all adjusted p < .01). Additionally, all groups performed better at age 6 than they did at age 4. However, the interaction effects were all insignificant (all p > .05). Results suggest that children who are DHH perform more poorly than their TH peers on these subtests of concept vocabulary knowledge at ages 4 and 6, and the magnitude of group differences does not change over time.

Finally, on the Quantity subtest, the full model revealed a significant main effect of time [F(1, 212) = 138.24, p < .001] and, unique to this subtest, a significant interaction effect between group and time [F(2, 212) = 3.05, p = .050]. Follow-up pairwise contrasts revealed that participants performed better on the subtest at age 6 than they did at age 4 (all adjusted p < .001) but that children who are DHH do not grow as quickly in their knowledge as children with TH. Thus, for this particular subtest, the data for children who are DHH deviate from the upward trend that their typical hearing peers demonstrate, suggesting that children with DHH present with particular difficulty in learning quantity concept words.

Research question 3

Research question 3 evaluated the influence of age of amplification and general vocabulary knowledge on overall BBCS:E test performance in children who are DHH only (i.e., HA and CI groups). Prior to conducting the multiple linear regression, assumptions were evaluated. There was independence of residuals as assessed by a Durbin–Watson statistic of 1.864 There was linearity as assessed by visual inspection of plot of studentized residuals versus unstandardized predicted values as well as the partial plots between each independent and dependent variable. There was homoscedasticity, as assessed by visual inspection of a plot of studentized residuals versus unstandardized predicted values. There was no multicollinearity, as assessed by Tolerance (all >0.90) There were no outlier and influential cases, as assessed by studentized deleted residual (all <3), leverage value (all <0.2), and Cook’s distance (all <1). The residuals are approximately normally distributed by visual inspection of P–P plot of regression standardized residual.

The multiple regression model statistically significantly predicted knowledge of concept vocabulary, as measured by BBCS:E test performance, F(2, 65) = 81.21, p < .001. The mean and standard deviation for the BBCS:E total composite score, EOWPVT at time 2, and age at amplification (measured in months) are 49.57 ± 16.84, 78.03 ± 17.64, and 13.99 ± 14.35, respectively. General vocabulary knowledge (measured by EOWPVT-4 raw score) statistically significantly predicted BBCS:E test performance, p < .001. Age at amplification did not statistically significantly predict BBCS:E test performance, p = .702. See Table 3 for regression coefficients and standard errors.

Discussion

The overall purpose of this study was to update the literature on concept vocabulary knowledge in children who are DHH, as all concept word knowledge studies with this population were conducted in the 1970s and 1980s (Bracken & Cato, 1986; Brenza et al., 1981; Davis, 1974). Because current hearing technology provides children who are DHH with better audibility than did technology from the 1980s (ASHA, n.d.; NIDCD, 2022) which impacts vocabulary outcomes (e.g., Davidson et al., 2014; Davis, 1974), it is theoretically possible that present-day children who are DHH could have different outcomes than those observed in the 1980s. The first research question asked whether present-day children who are DHH perform more poorly on measures of concept word knowledge than their typical hearing peers. Results reveal that, despite recent technological advances (ASHA, n.d.; NIDCD, 2022), children who are DHH still present with deficits in concept word knowledge.

More specifically, analyses revealed significant differences in performance between the TH and CI groups across all subtests, which included the School Readiness Composite and Direction/Position, Self-/Social Awareness, Texture/Material, Quantity, and Time/Sequence subtests. Additionally, analyses revealed a significant difference in performance between the TH and HA groups on the Direction/Position, Self-/Social Awareness, Texture/Material, Quantity, and Time Sequence subtests. Interestingly, the TH and HA groups did not perform differently on the School Readiness Composite, suggesting that children with HA may perform similarly to their TH peers on early developing concept words (e.g., colors, shapes). Thus, as a group, children with HA enter school performing similarly to their typical hearing peers on early, pre-academic concept vocabulary. Their concept vocabulary knowledge may appear as a relative strength in comparison to their peers with CI whose scores do significantly differ from TH peers, giving a false sense of proficiency in concept vocabulary knowledge in children with HA. However, the data suggest that children with HA do not maintain this level of performance across other types of concept knowledge, as their scores significantly differ from the TH group on the remaining subtests of the BBCS:E. These discrepancies have important clinical implications. Professionals must recognize that, although children with HA may perform similarly to developmental expectations on early concept vocabulary like shapes and colors, they likely still present with deficits on other concept words and may still benefit from support to learn these words.

Interestingly, this work found a performance difference between children who are DHH and their peers with TH on a norm-referenced measure of vocabulary knowledge. Other studies like Werfel and Douglas (2017) and Lund (2016) find that norm-referenced assessments may underestimate differences between children with TH and children who are DHH in functional settings. That is, children who are DHH and using CI or HA may perform in the range of normal according to test norms. However, when the performance of children who are DHH is compared to a sample of children with TH matched for characteristics beyond age (e.g., nonverbal cognition, socioeconomic status), performance differences between groups are larger (Lund, 2016). Additionally, when functional assessments that measure ecologically valid skills, like language samples, explore differences between DHH and TH children, differences are observed that may be obscured by norm-referenced measures (Werfel & Douglas, 2017). That a difference in performance exists on a norm-referenced measure for this study may indicate that functional assessment of concept knowledge could reveal even wider gaps between children who are DHH and children who are TH.

The second question asked whether children who are DHH learn concept vocabulary at the same rate as their typical hearing peers between the ages of 4 and 6. Results revealed that children in all groups performed better on the BBCS:E at age 6 than they did at age 4, suggesting growth over time for all groups. However, results for all subtests (except Quantity) lacked a significant interaction effect between group and time, suggesting that all three groups improved their concept word knowledge at the same rate between the ages of 4 and 6. Additionally, the significant interaction effect for the Quantity subtest revealed that children who are DHH learned quantity concept words at a slower rate than their TH peers.

Together, these preliminary data reveal a gap in concept vocabulary knowledge in children who are DHH and suggest that this gap persists to at least 6 years of age (i.e., the beginning of elementary school). The literature currently suggests that children acquire most of their concept vocabulary knowledge by age 6 (Boehm, 1991) and that children who are DHH show minimal growth in concept knowledge thereafter (Brenza et al., 1981). Thus, this gap may persist. A potentially persistent gap in concept vocabulary knowledge likely has consequences for children who are DHH. The literature suggests that concept vocabulary knowledge correlates with academic performance, specifically with readiness to learn (Panter, 2000; Piersel & McAndrews, 1982), achievement test outcomes (Breen, 1985; Piersel & McAndrews, 1982; Zucker & Riordan, 1990), and intelligence test outcomes (Howell & Bracken, 1992). We know that children who are DHH struggle academically (Breland et al., 2022; Zussino et al., 2022). It is possible that academic discrepancies could be, in part, exacerbated by concept vocabulary deficits.

Finally, the third research question asked about variables that predict concept word knowledge in children who are DHH. The EOWPVT-4, a measure of omnibus expressive vocabulary knowledge, is a strong predictor of concept vocabulary knowledge. This finding builds upon Breen’s (1985) results that reveal correlations between receptive vocabulary and measures of concept vocabulary. Together, these findings suggest that omnibus vocabulary knowledge and concept vocabulary knowledge likely go hand in hand. Thus, because omnibus vocabulary is such a strong predictor of concept vocabulary knowledge, deficits in omnibus vocabulary measures should raise concerns about concept vocabulary knowledge.

Age at amplification was not a significant predictor of concept vocabulary outcomes. It is true that children are receiving amplification younger than ever before (e.g., ASHA, 2020; Ching et al., 2017) and that early amplification positively impacts language outcomes (e.g., Wu et al., 2023). The fact that age of amplification did not predict BBCS:E outcomes in this study suggests that early amplification does not, alone, resolve the impact that hearing loss has on language development, specifically on concept vocabulary knowledge. Children who are DHH, even with early amplification, still experience language deficits, including deficits in concept vocabulary knowledge, which need attention and support to facilitate best-case scenario outcomes for children who are DHH. It may be, instead, that the benefits of early amplification are best harnessed with additional intervention support. Grey et al. (2022), for example, found that entering intervention by 6 months was a better predictor of spoken language outcomes than meeting the EDHI benchmark of amplification by 3 months. The combination of factors that best leads to vocabulary growth provides avenues for additional investigations.

Limitations and future directions

Some limitations with this study exist, which provide direction for future investigations. First, this study only analyzed concept vocabulary knowledge in children between the ages of four and six. Results reveal persistent deficits throughout these ages, but the trajectory of development past the age of 6 in children who are DHH with current hearing technology is unknown. Future studies should evaluate the trajectory of concept word knowledge past the age of 6 to determine whether children who are DHH improve concept vocabulary knowledge and reach levels of performance similar to that of their TH peers.

Second, this study speculates that observed deficits in concept vocabulary knowledge could impact the academic performance and experience of children who are DHH. However, this has not been directly studied with this population. Future studies should study the direct impact of concept vocabulary deficits on this population’s academic performance, ability to learn, and ability to follow directions in the classroom. This next step would reveal the extent that these deficits play a role in the classroom experience and performance of children who are DHH.

Third, this study only evaluated concept vocabulary knowledge in a single-word, expressive measure. This study does not inform this population’s ability to comprehend concept words and to use concept words in conversation, both of which provide important insight into their overall performance with this type of vocabulary. Future studies should investigate these clinically relevant topics to paint a comprehensive picture of concept vocabulary knowledge and skills in children who are DHH.

Conclusion

In conclusion, all existing studies on concept knowledge in children who are DHH tell the same story: This population presents with deficits in concept vocabulary knowledge. Current data suggests that these deficits persist to at least 6 years of age, even though children who are DHH demonstrate growth between ages 4 and 6. Notably, the progress of children who are DHH did not indicate that they are catching up with the knowledge of their same-age peers. Future studies should consider long-term developmental trajectories and impact on academic outcomes for this population. These findings underscore the need for targeted interventions to support concept vocabulary development in children who are DHH.

Funding

This work was supported by the National Institute on Deafness and Other Communication Disorders at National Institutes of Health [grant number R01 DC017173 to K.L.W./E.L.].

Conflicts of interest

None declared.

References