Parents Do Understand: Agreement Between Self- and Parent-Reported Psychosocial Adjustment in Adolescents With Cochlear Implants

Abstract

Objective:

The objective of this study was to examine self- and parent-reporting of social, emotional, and behavioral adjustment in adolescents with and without prelingual hearing loss and cochlear implants.

Method:

The self- and parent-completed Behavior Assessment System for Children was used to assess social, emotional, and behavioral adjustment in adolescents aged 12–19 years. Sixty-five adolescents and their parents were recruited from the Midwestern United States and participated in the present study. Analyses were conducted to (a) examine self-reported social, emotional, and behavioral adjustment in adolescents with cochlear implants (ACIs) as compared to adolescents with hearing (AH); (b) examine associations between self-reported adjustment and performance-based neurocognitive skills; and (c) determine the degree of agreement between self- and parent-reported adjustment in both samples of adolescents.

Results:

The sample of ACIs self-reported significantly higher levels of atypicality and depression, and significantly lower levels of interpersonal relations, self-reliance, and personal adjustment when compared to the AH sample. Self-reported adjustment and performance-based neurocognitive skills were associated in both groups of adolescents. Specifically, better language, verbal working memory, and inhibition–concentration skills were associated with fewer internalizing problems, fewer emotional symptoms, and stronger personal adjustment in the ACI sample and with fewer school problems in the AH sample. Finally, in the ACI sample, results revealed agreement between self- and parent-ratings of social, emotional, and behavioral adjustment. In the AH sample, however, results revealed divergence between self- and parent-ratings of anxiety and hyperactivity.

Conclusions:

Self-reported measures of social, emotional, and behavioral adjustment revealed differences between samples of ACI and AH. Notably, self- and parent-reporting showed agreement, supporting the use of parent-reporting as a valid measure of adjustment in clinical ACI samples.

The cascading effects of prelingual hearing loss and delayed access to language on child and adolescent neurocognitive (language and executive functioning [EF]) skills are well documented (Castellanos et al., 2016; Kronenberger, Beer, et al., 2014). Several recent reports have also identified domains of psychosocial adjustment, which encompasses social, emotional, and behavioral adaptation to challenges, which may be adversely affected by prelingual hearing loss (Bigler et al., 2019; Castellanos et al., 2018, 2020; Cejas et al., 2020). In order to help clinicians diagnose and treat individuals based on the specific nature of their social, emotional, and behavioral adaptation, the Diagnostic Statistical Manual of Mental Disorders, categorizes psychosocial adjustment into two primary categories of behavioral problems: internalizing and externalizing (Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision; American Psychiatric Association, 2022). Internalizing behavioral problems center on symptoms of internal emotional states such as withdrawal, anxiety, and depression; whereas externalizing behavioral problems include acting-out and disruptive external behaviors, such as aggression, conduct problems, and hyperactivity.

Psychosocial adjustment is one component of health-related quality of life, which is a broader construct that refers to overall well-being and satisfaction with life in response to health conditions or treatments (Cejas et al., 2023; Warner-Czyz et al., 2022). Psychosocial adjustment and quality of life both encompass psychological health and social relationships, while quality of life covers a wider range of factors contributing to an individual's well-being such as of physical health, functional independence, spirituality, and environmental quality. Research employing quality-of-life measures have sought to examine the degree to which a child's well-being and life satisfaction improved following cochlear implantation (see Warner-Czyz et al., 2022, for a comprehensive review). We focus our attention on measures of psychosocial adjustment in adolescents with and without prelingual hearing loss because these targeted measures provide valuable insights into at-risk or diagnosable behavioral problems, which can be effectively addressed and modified through individualized treatment and intervention.

Associations Between Psychosocial and Neurocognitive Functioning

While variability in child/adolescent hearing, demographic, communication factors, and schooling placement affect reporting, in general, findings from proxy- and self-reporting suggest that children and adolescents who are deaf or hard of hearing have difficulties in several, but not all aspects of psychosocial adjustment. The prevalence rate of self-reported difficulty in domains of psychosocial adjustment ranges from 25% to 54% in adolescents who are deaf or hard of hearing (Cejas et al., 2020; Fellinger et al., 2008; Wallis et al., 2004). Adolescents who are deaf or hard of hearing self-report a higher incidence of internalizing and externalizing problems (Cejas et al., 2020; van Eldik, 2005), social exclusion (Kouwenberg et al., 2012; Warner-Czyz et al., 2018), and coercion during the school year as compared to adolescents with hearing (AH; Warner-Czyz et al., 2018). This pattern is of concern, particularly during the adolescent period of development, because studies on AH samples report associations between poorer psychosocial adjustment and school absenteeism (Finning et al., 2019), and poorer psychosocial adjustment and perpetrating or experiencing bullying behaviors in school (Nansel et al., 2001, 2004). Similarly, in adolescents with hearing aids (HAs) or cochlear implants (CIs), experiencing higher levels of bullying in school are associated with poorer self-reported psychosocial adjustment, including heightened feelings of sadness and anger (Kouwenberg et al., 2012).

A large portion of the individual variability in pediatric HA/CI users' psychosocial adjustment remains unexplained even after accounting for hearing and demographic factors. In fact, the individual variability in psychosocial adjustment within groups of pediatric HA/CI users appears to be greater than the group variability between pediatric HA/CI users and their hearing peers (Castellanos et al., 2018). In recent years, researchers have moved away from describing “between group” variability in favor of uncovering basic underlying information processing skills that may help to predict “within-group” variability in pediatric HA/CI users' psychosocial adjustment. In order to explain this “within-group” variability, a growing number of studies have investigated possible associations between psychosocial adjustment in pediatric HA/CI users and aspects of their neurocognition including nonverbal intelligence, speech-language, and EF skills.

Language is a critical tool for developing and managing foundational beliefs about the self and may be crucially important for the development of adaptive psychosocial skills (Alderson-Day & Fernyhough, 2015; Winsler et al., 2009). Brown and Cornes (2015) illustrated the association between spoken language proficiency and self-reported psychosocial adjustment in adolescents with HAs or CIs. Consistent with previous findings, between-group analyses found that adolescents with HAs or CIs self-reported poorer psychosocial adjustment, such as experiencing greater internalizing and externalizing behavioral problems, somatic complaints, and greater social problems, compared to their AH peers (Brown & Cornes, 2015). Interestingly, within-group analyses demonstrated that these self-reported difficulties were strongly associated with language use in the home such that adolescents with HAs or CIs who use Signed English or Sign Language in the home self-reported more difficulties with psychosocial adjustment than their peers with HAs or CIs who used spoken English. This finding aligns with previous research suggesting that the shared use of a language may be associated with easier parent-adolescent communication and improved psychosocial adjustment in adolescents, though the directionality of this relationship remains unclear (van Eldik et al., 2004; Wallis et al., 2004).

Measures of speech intelligibility, which index the robustness of lexical representations in speech, also help to explain variability in psychosocial adjustment in samples of adolescents with cochlear implants (ACIs). Freeman et al. (2017) reported associations between CI users' speech intelligibility—that is, how well their speech is understood by inexperienced listeners—and parent-reported psychosocial adjustment. Specifically, CI users aged 7–19 years with poorer speech intelligibility scores were rated by their parents as displaying higher prevalence rates of depression, attention problems, atypicality, and withdrawal. Relatedly, ACIs with higher speech perception, speech intelligibility, and reading scores during elementary school were rated by their parents during adolescence as displaying average or better levels of assertive and cooperative behaviors (Moog et al., 2011). Poorer perception and understanding of speech in noise have also been associated with poorer parent-reported psychosocial adjustment in ACIs (Huber, Burger, et al., 2015).

Associations between psychosocial adjustment and EF (self-regulation of thought, behavior, and emotion) have been extensively studied in children and AH. Because EF skills are recruited to regulate behavior and emotions, deficits in EF skills are associated with behavioral and emotional problems, such as psychiatric diagnoses of attention-deficit/hyperactivity disorder (Barkley, 2012). Indeed, in samples of AH, EF skills (working memory, inhibition–concentration, and attention shifting) obtained during the preschool period of development are associated with self-reported externalizing behavioral problems during adolescence (Fleming et al., 2020). In children and adolescents who are deaf or hard of hearing aged 5–18 years, teacher-completed questionnaire measures of language competence and EF skills were associated with perceived psychosocial skills, such that deaf or hard of hearing children and adolescents with higher language competence and EF skills (e.g., inhibition, working memory) were rated as displaying fewer emotional symptoms, conduct problems, hyperactivity-inattention, peer problems, and higher prosocial behavior than their peers (Hintermair, 2013). Consistent with these findings, pediatric CI users appear to be at risk for poorer adjustment in the psychosocial domains of hyperactivity–impulsivity and oppositional behavior related to language, working memory, and inhibition–concentration skills (Castellanos et al., 2018). Taken together, these studies highlight the cascading effects of language and EF skills on psychosocial functioning in children and adolescents, particularly those who are deaf or hard of hearing.

Parent–Proxy Reporting and Agreement

The vast majority of research findings on the effects of prelingual hearing loss on psychosocial adjustment have come from studies using proxy (observer)–completed questionnaires (although with some exceptions see Warner-Czyz et al., 2022), partly because (a) self-reported measures used in the diagnostic testing of psychosocial adjustment have been validated for use only in older school-aged children and adolescents since young children often lack the cognitive and linguistic skills necessary to complete questionnaires (Castellanos et al., 2020), and (b) limitations exist in the availability of validated questionnaires that assess the same psychosocial domains across both self- and proxy-reporting. Similar to self-reported rates, parent-reported prevalence rates of clinically significant internalizing and externalizing behavioral problems in adolescents who are deaf or hard of hearing are estimated at 35%–41% (Castellanos et al., 2018; Van Eldik et al., 2004). Parent-reported data collected by van Eldik et al. (2004) demonstrate that children and adolescents who are deaf or hard of hearing experience more psychosocial difficulties across developmental time; specifically, parents reported more symptoms of depression, anxiety, and social problems in their adolescents aged 12–18 years than in their school-age children aged 4–11 years.

Although questionnaires using parent–proxy ratings have been extensively validated (e.g., Levy et al., 2017; Reynolds & Kamphaus, 2004) and previous studies have shown that parent–completed questionnaires have good predictive validity of CI users' psychosocial functioning (Snyder et al., 2004), factors such as parent experiences, expectations, bias, and awareness of behaviors can affect parent ratings. Indeed, a meta-analysis on school-age children and adolescents with and without chronic illnesses indicates that parental expectations affect parents' ratings of social, emotional, and physical skills (Eiser & Morse, 2001a, 2001b), and that self–parent agreement is higher for more observable physical attributes than for less observable emotional and internal states (Eiser & Morse, 2001a, 2001b). Moreover, we speculate that the risk of parent biases affecting parent ratings may be particularly high when adolescents are embedded in different cultures or school settings with expectations for behaviors that differ from the lived experiences of the parent. To address potential parental biases, several studies have collected self- and parent-reported data on psychosocial adjustment in samples of ACI, though few studies have collected these data using the same questionnaire for both informants.

Self- and proxy-reported data on psychosocial adjustment afford a more comprehensive understanding of outcomes following cochlear implantation because adolescents and parents each provides unique, varied, and rich sources of information. As such, we advocate for a multimethod–multisource approach to assessment in which multiple informants report across multiple settings such as home and school (Holmbeck et al., 2002). Multiple perspectives about the sensitivity to, awareness of, and tolerance of a behavioral problem are gained through a multimethod–multisource approach. One such study found agreement among self- and parent-proxy–reporting in several domains of psychosocial adjustment in ACIs with and without additional handicaps as compared to their AH peers (Huber, Burger, et al., 2015). The authors reported high self-parent agreement in the ACI sample aged 12–17 years for all domains assessed: emotional symptoms, hyperactivity-inattention, conduct problems, peer-problems, and prosocial behavior. In contrast, self-parent agreement was only significant for two domains (emotional symptoms and prosocial behavior) in the sample of AH peers. Interestingly, self-parent agreement was significantly higher in the ACI sample than in the sample of AH peers.

In a later study, Huber, Pletzer, et al. (2015) examined psychosocial differences among ACIs aged 12–17 years as a function of school placement using self- and parent-proxy–reporting on the Strength and Difficulties Questionnaire (Huber, Pletzer, et al., 2015). Although informant agreement was not a stated goal of the study, the authors reported that ACIs in special schools for the deaf or hard of hearing displayed more conduct problems than ACIs in mainstream schools, a pattern that was replicated across all informants. Finally, informant agreement appears to improve with age, though the reasons underlying these effects remain unknown. Using the KINDL questionnaire, Huber (2005) reported high self-parent agreement on measures of quality of life (physical and psychosocial adjustment) in ACIs aged 13–16 years, but low self-parent agreement in school-age CI users aged 8–12 years. Possible explanations for the results across these two questionnaires include that adolescents, compared to school-age children, may be better at identifying and communicating their well-being and/or more adept at completing behavioral rating scales.

The present study examined three primary questions concerning psychosocial adjustment in prelingually deaf ACIs: (a) How does the self-reported psychosocial adjustment of ACIs compare to that of AH peers? (b) What are the associations between self-reported psychosocial adjustment and measures of neurocognitive (language, verbal working memory, visual–spatial working memory, fluency–speed, inhibition–concentration) functioning in ACIs, compared to AH peers? (c) How well do self- and parent-reports of psychosocial adjustment agree in samples of ACIs compared to AH peers? To date, no study has employed the same questionnaire (assessing the same domains) to examine the potential differences in CI users' self- versus parent-reported psychosocial adjustment.

Method

Participants

The data presented in this article were collected as part of a larger study examining the neurocognitive, speech-language, and psychosocial outcomes of pediatric CI users (Castellanos et al., 2018). Sixty-five adolescents and their parents were recruited from the Midwestern United States and participated in the present study. To be included in the present study, adolescent participants were required to be at aged 12 and 19 years and eligible to complete the adolescent version of the Behavior Assessment System for Children–Second Edition (BASC-2).

Adolescents With Cochlear Implants (n = 31)

Adolescents with cochlear implants (ACIs) were recruited from a large university hospital–based CI clinic and local advertisements. ACIs met the following inclusionary criteria: (a) prelingual severe-to-profound sensorineural hearing loss (> 70 dB HL in the better hearing ear prior to age 2 years), (b) received their CI prior to age 7 years, (c) used their CI for 7 years or more, (d) used a currently available state-of-the-art multichannel CI system, (e) lived in a household with spoken English as the primary language, and (f) passed a screening performed by licensed speech-language pathologists prior to testing, confirming no additional developmental, neurological, or cognitive conditions were present other than hearing loss. Demographics and hearing history variables obtained for the ACI sample are provided in Table 1. At testing, 28 (90%) ACIs reported using oral communication, while three (10%) ACIs reported using simultaneous communication strategies. Etiology of deafness included unknown (n = 20, 64.5%), familial (at least one immediate family member also had deafness of unknown etiology, n = 6, 19.4%), meningitis (n = 2, 6.5%), Mondini malformation (n = 1, 3.2%), auditory neuropathy (n = 1, 3.2%), and large vestibular aqueduct (n = 1, 3.2%).

Table 1.

Participant demographics and hearing history.

Variable	Hearing group
	ACI		AH
n	31		34
	M (SD)	Range	M (SD)	Range
Onset of deafness (mos.)	2.35 (6.30)	0.00–24.00	—
Age at implantation (mos.)	37.87 (18.14)	17.25–75.76	—
Age at testing (yrs.)	15.11 (2.30)	12.08–19.13	14.71 (1.87)	12.07–19.80
Duration of CI use (yrs.)	11.96 (1.99)	8.18–15.70	—
Pre-implant PTAa	105.42 (12.08)	85.00–118.43	—
CMRSb	4.77 (0.72)	2.00–5.00	—
Nonverbal IQc	53.48 (7.08)	35.00–65.00	54.21 (6.48)	38.00–66.00
	Count (% of sample)
Hearing deviced			—
Bilateral CIs	10 (32.3%)
Unilateral CI	19 (61.3%)
CI and contralateral HA	2 (6.4%)
Etiology of hearing loss			—
Meningitis	2 (6.5%)
Other/unknown	29 (93.5%)
Sex
Female	17 (54.8%)		17 (50.0%)
Male	14 (45.2%)		17 (50.0%)
Race
Black	—		6 (17.6%)
Multiracial	1 (3.2%)		—
White	30 (96.8%)		28 (82.4%)
Ethnicity
Hispanic	1 (3.2%)		1 (2.9%)
Not Hispanic	30 (96.8%)		33 (97.1%)
Income level
< $5,000	—		—
$5,000–$9,999	—		—
$10,000–$14,999	1 (3.2%)		—
$15,000–$24,999	1 (3.2%)		2 (5.9%)
$25,000–$34,999	3 (9.7%)		1 (2.9%)
$35,000–$49,999	5 (16.1%)		8 (23.5%)
$50,000–$64,999	3 (9.7%)		5 (14.7%)
$65,000–$79,999	3 (9.7%)		3 (8.8%)
$80,000–$94,999	5 (16.1%)		1 (2.9%)
> $95,000	7 (22.6%)		13 (38.2%)

Note. Em dashes indicate data not available. ACI = adolescent with cochlear implant; AH = adolescent with hearing; mos. = months; yrs. = years.

^aPre-implant unaided pure-tone average (PTA) in the better ear for the frequencies 500, 1000, and 2000 Hz in dB HL.

^bCommunication Mode Rating Scale (CMRS) at testing (coded on a 1 [mostly sign] to 6 [auditory verbal] scale, with values of 1–3 reflecting simultaneous communication strategies [sign and speech to varying degrees of emphasis] and 4–6 reflecting oral communication strategies [speech used exclusively with no formal sign language other than gestures]; Geers & Brenner, 2003).

^cWechsler Abbreviated Scale of Intelligence Matrix Reasoning subtest T score for nonverbal intelligence.

^dCI = cochlear implant; HA = hearing aid.

Adolescents With Normal Hearing (n = 34)

Adolescents with normal hearing (AH) were recruited from advertisements placed in the community. The AH control sample passed a basic audiometric hearing screening (headphones were used to test each ear individually at frequencies of 500, 1000, 2000, and 4000 Hz at 20 dB HL), and reported no significant developmental, neurological, or cognitive delays. Characteristics of the AH sample are also summarized in Table 1.

Procedure

All study procedures were reviewed and approved by the local institutional review board (Protocol Number: 1011003908), and written informed consent was obtained for all participants or parents prior to initiation of study procedures. Licensed speech-language pathologists evaluated the ACI sample and administered the language tests in the participant's mode of communication used at school. Speech-language pathologists or experienced psychometric technicians also evaluated the AH sample. A licensed clinical psychologist (W.K.G) supervised all examiners.

Measures

Psychosocial Adjustment

The self- and parent-completed BASC-2 (Reynolds & Kamphaus, 2004) for the adolescent age range (ages 12–21 years) was used to assess psychosocial (social, emotional, and behavioral) adjustment. In the present study, BASC-2 proxy reports were completed by 53 mothers, nine fathers, and three grandparents. Reynolds and Kamphaus (2004) report strong psychometrics for the self- and parent-completed BASC-2, including construct validity (the r values range from .65 to .78 for self-completed scales; the r values range from .59 to .72 for parent-completed scales¹), internal consistency (Cronbach's alpha values range from .87 to .93 for self- and parent-completed BASC-2 forms), test–retest reliability (r values range from .82 to .85 for self- and parent-completed BASC-2 forms), and interrater reliability among parents of the same child (r values range from .82 to .84).

The self-completed BASC-2 contains 176 items in the form of statements that young people may use to describe how they think, feel, or act. Participants are instructed to endorse the first set of statements as either “true” or “false,” and the remaining statements as either “never,” “sometimes,” “often,” or “almost always.” Items constitute 12 clinical subscales and four adaptive subscales, which are aggregated to create five indices of behavior (see Table 2 for brief descriptions of subscales). Six clinical subscales are present in both the self- and parent-report versions, which include atypicality, anxiety, depression, somatization, attention problems, and hyperactivity. Only these six clinical subscales were used in the present study to compare agreement across self- and parent-reporting. The BASC-2 uses age–sex norms and scores are expressed as T scores (M = 50, SD = 10). Higher T scores on clinical subscales/indices indicate greater problems with psychosocial adjustment (at-risk: T score range: 60–69; clinically significant: T scores ≥ 70), while lower T scores on adaptive subscales/indices indicate greater problems with adaptive behavior (at-risk: T score range: 31–40; clinically significant: T scores ≤ 30).

Table 2.

Brief descriptions of Behavior Assessment System for Children–Second Edition (BASC-2) subscales and indices.

BASC-2 subscales and indices	Brief description
Clinical
Attitude to school	Negative feelings like dissatisfaction with school
Attitude to teachers	Negative feelings like resentment toward teachers
Sensation seeking	Tendency to engage in risk-taking behaviors
Atypicalitya	Thoughts and behaviors that are odd or unusual
Locus of control	Belief that external factors dictate life events and outcomes beyond one's control
Social stress	Negative feelings like stress and tension in personal relationships
Anxietya	Negative feelings like worry and fear
Depressiona	Negative feelings like sadness, hopelessness, and a lack of interest in activities
Sense of inadequacy	Negative feelings like being unsuccessful and unable to achieve goals
Somatizationa	Tendency to complain about relatively minor physical problems and discomforts
Attention problemsa	Difficulty maintaining focus and concentration
Hyperactivitya	Tendency to have excessive energy, impulsivity, and difficulty sitting still
Adaptive
Relations with parents	Positive feelings like esteem toward parents and feelings of support
Interpersonal relations	Sense of having good social relationships and ability to interact in social settings
Self-esteem	Sense of worth and value as a person
Self-reliance	Sense of confidence in one's ability to solve problems and make decisions without support from others
Indices
School problems	Attitude to school, attitude to teachers, and sensation seeking
Internalizing problems	Atypicality, locus of control, social stress, anxiety, depression, sense of inadequacy, and somatization
Inattention/hyperactivity	Attention problems and hyperactivity
Emotional symptoms	Social stress, anxiety, depression, sense of inadequacy, self-esteem, and self-reliance
Personal adjustment	Relations with parents, interpersonal relations, self-esteem, and self-reliance

^aThe six clinical subscales that are present in both the self- and parent-report versions of the BASC-2. Indices are composite of the subscales listed.

Neurocognitive Tests

Gold standard neurocognitive tests with strong psychometrics including internal consistency, test–retest reliability, and construct validity were administered in their standardized format using spoken instructions when appropriate (see Castellanos et al., 2018, for a more detailed description of specific tests). T scores from the Matrix Reasoning subtest of the Wechsler Abbreviated Scale of Intelligence (Wechsler, 1999) was used to assess global nonverbal intelligence. Additionally, composite scores² were created from individual neurocognitive assessments: (a) language, (b) verbal working memory, (c) visual–spatial working memory, (d) fluency–speed, and (e) inhibition–concentration.

(a) Language: standard scores on the Peabody Picture Vocabulary Test–Fourth Edition (Dunn & Dunn, 2007) and Core Language standard scores from the Clinical Evaluation of Language Fundamentals–Fourth Edition (Semel et al., 2003) were used to assess language skills. For CI users using total communication (N = 3, 9.7%), Signed Exact English accompanied the administration of the language tests. (b) Verbal working memory: scaled scores on the Digit Span Forward and Backward subtests of the Wechsler Intelligence Scale for Children–Third Edition (Wechsler, 1991) and scaled scores on the Visual Digit Span subtest of the Wechsler Intelligence Scale for Children–Fourth Edition–Integrated (WISC-IV-I; Wechsler et al., 2004) were used to assess verbal working memory capacity. (c) Visual–spatial working memory: scaled scores on the Spatial Span Forward and Backward subtests of the WISC-IV-I (Wechsler et al., 2004) were used to assess Visual–Spatial Working Memory capacity. (d) Fluency–speed: standard scores on the Pair Cancellation subtest of the Woodcock–Johnson Tests of Cognitive Abilities–Third Edition (Woodcock et al., 2001), and scaled scores on the Coding and Coding Copy subtests of the WISC-IV-I (Wechsler et al., 2004) were used to assess fluency–speed skills. (e) Inhibition–concentration: omissions, commissions, and response time variability standard scores from the Test of Variables of Attention (Leark et al., 1996) were used to assess inhibition–concentration skills.

Data Analysis

First, two-tailed independent samples t tests were used to compare the two samples of adolescents on their self-reported adjustment. To provide a comprehensive understanding of their social, emotional, and behavioral adjustment, self-reported BASC-2 clinical and adaptative subscale scores along with self-reported index scores were analyzed. Cohen's d was used to quantify small (d = 0.20), medium (d = 0.50), or large (d = 0.80) effect sizes. Adolescents were classified as at-risk for poorer self-reported adjustment if their BASC-2 T scores fell within at-risk to clinically significant range. Fisher's exact tests were then used to determine if there was a significant association between sample (ACI, AH) and being classified as at-risk for poorer self-reported adjustment. To assess how our sample of adolescents compared to large-scale studies reporting sex differences in adjustment, a multivariate analysis of variance (MANOVA) with self-reported subscale/index scores as dependent variables and sex and sample (ACI, AH) as between-subjects factors was conducted to examine the effect of sex and hearing status on psychosocial outcomes. Significant interactions between sex and sample were examined with planned pairwise comparisons.

Next, composite scores for language, verbal working memory, visual–spatial working memory, fluency–speed, and inhibition–concentration were computed by summing z-transformed scores derived from the means and standard deviations of the current sample of adolescents' performance on the individual neurocognitive assessments (see Kronenberger, Colson, et al., 2014, for support of this technique). Composite scores were then used for all correlational analyses. In order to examine the relations between neurocognitive skills and psychosocial outcomes, composite scores were correlated with index scores from the BASC-2, while statistically controlling for nonverbal intelligence. One-tailed testing was used to examine our directional hypothesis that better neurocognitive skills are correlated with fewer behavior problems and better adaptive behaviors, while also reducing Type I error rate. To further reduce the risk of Type I error in the correlation coefficient, we opted to focus on the five BASC-2 indices rather than all 16 clinical and adaptive subscales. These partial correlation analyses, controlling for nonverbal intelligence, were carried out separately for the ACI and AH groups. Correlations were defined as weak (r = .10), moderate (r = .30), or strong (r = .50).

Finally, the six BASC-2 subscales common across both self- and parent-reports were analyzed to determine the degree of agreement between self- and parent-reported adjustment. A series of MANOVAs were conducted with reporter (self, parent) as the within-subjects factor and sample (ACI, AH) as the between-subjects factor on each of these six subscales. Partial eta-squared was used to quantify small (η_p² = .01), medium (η_p² = .06), or large (η_p² = .14) effect sizes. Significant interactions between reporter and sample were explored with pairwise comparisons. Correlational analyses were then carried out to evaluate the magnitude of self–parent agreement for each of the six shared subscales.

Results

Self-Reported Adjustment

ACI and AH samples did not differ on nonverbal IQ scores, t(63) = −0.43, p = .67; age, t(63) = 0.78, p = .44; sex (p = .44 by Fisher's exact test); or family income, t(59) = −0.53, p = .60; see Table 1. Self-reported BASC-2 T scores for the ACI and AH samples are shown in Figure 1, Panel A. Compared to AH controls, the ACI sample self-reported significantly poorer functioning across five subscales. Specifically, within the clinical subscales, the ACI sample self-reported significantly higher levels of atypicality, t(63) = 2.66, p = .01, d = 0.66, and depression, t(63) = 2.08, p = .04, d = 0.52, indicating greater problems with behavioral adjustment. Within the adaptive subscales, the ACI sample self-reported significantly lower levels of interpersonal relations, t(63) = −2.86, p = .01, d = −0.71; self-reliance, t(63) = −3.13, p < .01, d = −0.77; and personal adjustment, t(63) = −2.71, p = .01, d = −0.67, indicating greater problems with adaptive behavior.

Figure 1.

Mean self-reported scales for adolescents with cochlear implants (ACIs) and adolescents with hearing (AH). Panel A: Higher T scores on scales to the left of the vertical dashed line indicate greater problems in behavioral adjustment. Lower T scores on scales to the right of the vertical dashed line indicate greater problems in adaptive behavior. Only scales that appear in both the self-report and parent–report are depicted in Panel B. BASC-2 = Behavior Assessment System for Children–Second Edition. *p < .05; **p ≤ .01; ***p < .001.

Based on self-reports, the ACI sample was more likely than their AH peers to display at-risk to clinically significant delays across four subscales (all by Fisher's exact tests; see Table 3). Specifically, within the clinical subscales, the ACI sample was more likely than their AH peers to report at-risk to clinically significant delays in atypicality (p = .02), and social stress (p = .03). Within the adaptive subscales, the ACI sample was more likely than their AH peers to report at-risk to clinically significant delays in interpersonal relations (p = .04), and self-reliance (p = .03). Interestingly, self-reported BASC-2 T scores for the ACI sample tended to be close to the normative M of 50, whereas the AH sample tended to report scores in the better-adjusted direction relative to the normative M of 50.

Table 3.

Percent of self-reported elevated psychosocial scores.

BASC-2 subscales	ACI		AH		p value
	M (SD)	Percentage of elevated scores	M (SD)	Percentage of elevated scores
Clinical
Attitude to school	48.35 (13.07)	29.03	46.55 (9.56)	12.12	.09
Attitude to teachers	46.23 (8.61)	9.68	45.88 (9.84)	12.12	.54
Sensation seeking	52.23 (11.67)	32.26	47.62 (10.75)	14.71	.08
Atypicality	52.90 (11.13)	22.58	47.03 (6.18)	2.94	.02
Locus of control	50.26 (11.40)	16.13	46.56 (8.06)	11.76	.44
Social stress	49.71 (11.82)	25.81	45.68 (8.25)	5.88	.03
Anxiety	48.68 (8.93)	12.90	50.74 (11.18)	26.47	.15
Depression	48.84 (9.13)	12.90	44.76 (6.55)	2.94	.15
Sense of inadequacy	49.00 (11.19)	16.13	46.71 (10.35)	11.76	.44
Somatization	51.26 (11.72)	19.35	47.47 (10.07)	14.71	.43
Attention problems	49.48 (12.64)	22.58	47.12 (10.41)	8.82	.12
Hyperactivity	46.84 (10.70)	9.68	51.68 (12.67)	23.53	.12
Adaptive
Relations with parents	51.13 (8.96)	16.13	54.85 (7.50)	2.94	.08
Interpersonal relations	48.23 (10.20)	19.35	54.59 (7.64)	2.94	.04
Self-esteem	53.06 (8.43)	6.45	53.62 (7.27)	8.82	.55
Self-reliance	48.23 (9.78)	25.81	55.74 (9.56)	5.88	.03

Note. Percentages represent adolescents with BASC-2 T scores within the at-risk to clinically significant range. Higher T scores on clinical subscales indicate greater problems with psychosocial adjustment (at-risk: T score range: 60–69; clinically significant: T scores ≥ 70), while lower T scores on adaptive subscales indicate greater problems with adaptive behavior (at-risk: T-score range: 31–40; clinically significant: T scores ≤ 30). p values based on Fisher's exact tests, with bolding representing subscales in which p value < .05. To calculate the percentage of scores within the average range, subtract the percentage of elevated scores from 100%. ACI = adolescent with cochlear implant; AH = adolescent with hearing; BASC-2 = Behavior Assessment System for Children–Second Edition.

Sex Differences

A MANOVA revealed significant interactions between sex and sample for the relations with parents, F(1, 60) = 5.92, p = .02, η_p² = 0.09, and self-esteem, F(1, 60) = 4.82, p = .03, η_p² = 0.07, subscales, and the Personal Adjustment Index, F(1, 60) = 5.27, p = .03, η_p² = 0.08. Pairwise comparisons indicated that the female ACI sample self-reported significantly poorer relations with parents (female ACI M = 49.00, female AH peers M = 57.12; p < .01) and personal adjustment (female ACI M = 47.88, female AH peers M = 58.47; p < .01) than their female AH peers. Additionally, the female ACI sample self-reported significantly poorer self-esteem than the male ACI sample (female ACI M = 50.12, male ACI M = 56.64; p = .02).

Associations Between Self-Reported Adjustment and Neurocognition

Partial correlations between self-reported BASC-2 indices and neurocognitive composite scores, controlling for nonverbal intelligence, are shown in Table 4. Better fluency–speed skills were correlated with significantly stronger personal adjustment skills in both the ACI and AH samples. In contrast, visual–spatial working memory skills were not correlated with self-reported adjustment in either sample of adolescents.

Table 4.

Correlations between self-reported Behavior Assessment System for Children–Second Edition (BASC-2) indices and neurocognitive composite scores for adolescents with cochlear implants (ACI) and adolescents with hearing (AH), controlling for nonverbal intelligence.

BASC-2 indices	Neurocognitive composite scores
	Language		Verbal working memory		Visual–spatial working memory		Fluency–speed		Inhibition–concentration
	ACI	AH	ACI	AH	ACI	AH	ACI	AH	ACI	AH
df	28	31	28	31	28	31	28	31	28	31
School problems	−.01	−.32*	.01	−.35*	.08	−.05	.10	−.09	−.34*	−.34*
Internalizing problems	−.37*	−.06	−.38*	−.18	−.02	−.15	−.25	−.23	−.46**	−.03
Inattention/hyperactivity	−.05	−.17	−.15	−.26	−.05	−.14	−.22	.05	−.52**	−.32*
Emotional symptoms	−.38*	.00	−.41**	−.13	.03	−.09	−.26	−.28	−.44**	.08
Personal adjustment	.39*	−.09	.32*	.12	−.14	.06	.33*	.35*	.39*	−.01

Note. Lower T scores on the school problems, internalizing problems, inattention/hyperactivity, and emotional symptoms indices indicate fewer behavior problems. Higher T scores on the Personal Adjustment Index indicate better adaptive behaviors.

^*p < .05 based on one-tailed tests.

^**p ≤ .01.

In the ACI sample, better inhibition–concentration skills were significantly associated with all five BASC-2 indices, suggesting better adjustment. Moreover, better language skills and better verbal working memory skills were independently correlated with significantly fewer internalizing problems, fewer emotional symptoms, and stronger personal adjustment skills in the ACI sample. In the AH sample, better inhibition–concentration skills were associated with significantly fewer school problems and fewer symptoms of inattention/hyperactivity. Lastly, better language skills and better verbal working memory skills were independently correlated with significantly fewer school problems in the AH sample. Taken together, ACIs with better language and verbal working memory skills reported better emotional and interpersonal adjustment, whereas AH peers with better language and verbal working memory skills reported fewer school problems.

Parent-Reported Adjustment

Figure 1, Panel B presents the BASC-2 parent-reported subscales that differed significantly between the ACI and AH samples, focusing only on the six subscales that appear on both the self- and parent-reports of the BASC-2. The ACI sample were rated by their parents as experiencing significantly poorer adjustment on four out of the six (67%) subscales, with results revealing large magnitude group differences between the ACI and AH samples. Specifically, parents of the ACI sample, as compared to parents of AH peers, reported significantly more symptoms of atypicality, t(63) = 4.03, p < .001, d = 1.00; depression, t(63) = 2.48, p = .02, d = 0.62; attention problems, t(63) = 3.24, p < .01, d = 0.81; and hyperactivity, t(63) = 2.89, p = .01, d = 0.72, in their adolescent children. Based on parent-reports, the ACI sample was not more likely than their AH peers to display at-risk to clinically significant delays in the six clinical subscales assessed (see Table 5). Similar to the self-reported data, mean parent-reported scores for the ACI sample fell near the normative M of 50, whereas mean parent-reported scores for the AH sample fell well below the normative M of 50.

Table 5.

Percent of parent-reported elevated psychosocial scores.

BASC-2 subscales	ACI		AH		p value
	M (SD)	Percentage of elevated scores	M (SD)	Percentage of elevated scores
Anxiety	51.68 (11.27)	19.35	46.24 (11.24)	8.82	.19
Attention Problems	50.71 (9.75)	16.13	43.62 (7.85)	2.94	.08
Atypicality	51.58 (8.05)	12.90	44.85 (5.24)	2.94	.15
Depression	51.87 (10.63)	25.81	45.91 (8.73)	8.82	.07
Hyperactivity	50.55 (10.45)	16.13	44.21 (7.06)	2.94	.08
Somatization	47.45 (7.09)	6.45	44.88 (6.91)	5.88	.66

Note. Percentages represent adolescents, based on parent-reporting, with BASC-2 T scores within the at-risk to clinically significant range. Higher T scores indicate greater problems with psychosocial adjustment (at-risk: T-score range: 60–69; clinically significant: T scores ≥ 70). p values based on Fisher's exact tests. To calculate the percentage of scores within the average range, subtract the percentage of elevated scores from 100%. ACI = adolescent with cochlear implant; AH = adolescent with hearing; BASC-2 = Behavior Assessment System for Children–Second Edition.

Agreement Between Self- and Parent-Reporting

Table 6 presents descriptive statistics for self- and parent-reported adjustment. A MANOVA revealed significant interactions between reporter (self, parent) and sample (ACI, AH) for the anxiety, F(1, 63) = 6.16, p = .02, η_p² = 0.09, and hyperactivity, F(1, 63) = 12.76, p = .001, η_p² = 0.17, subscales, suggesting differences in mean subscale scores between self- and parent-reports as a function of the sample. Pairwise comparisons were performed to understand the nature of the significant Reporter × Sample interactions. In the ACI sample, pairwise comparisons indicated agreements between self- and parent-reported scores across all clinical scales. In the AH sample, however, pairwise comparisons indicated disagreements between self- and parent-reported scores such that AH controls self-reported significantly higher symptoms of anxiety (self M = 50.74, parent M = 46.24; p = .04) and hyperactivity (self M = 51.68, parent M = 44.21; p = .001) than was reported by their parents (see Figure 2).

Table 6.

Descriptive statistics for self- and parent-reported Behavior Assessment System for Children–Second Edition (BASC-2) scores.

BASC-2 subscales	ACI			AH
	Self-reported	Parent-reported	Self–parent agreement	Self-peported	Parent-reported	Self–parent agreement
	M (SD)	M (SD)	r	M (SD)	M (SD)	r
Anxiety	48.68 (8.93)	51.68 (11.27)	.24	50.74 (11.18)	46.24 (11.24)	.45a*
Attention Problems	49.48 (12.64)	50.71 (9.75)	.65**	47.12 (10.41)	43.62 (7.85)	.32*
Atypicality	52.90 (11.13)	51.58 (8.05)	.28	47.03 (6.18)	44.85 (5.24)	.33*
Depression	48.84 (9.13)	51.87 (10.63)	.42*	44.76 (6.55)	45.91 (8.73)	.27
Hyperactivity	46.84 (10.70)	50.55 (10.45)	.50**	51.68 (12.67)	44.21 (7.06)	.05
Somatization	51.26 (11.72)	47.45 (7.09)	.23	47.47 (10.07)	44.88 (6.91)	.46**

Note. Only the six subscales shared between the self and parent versions of the BASC-2 were analyzed. BASC-2 subscale scores are expressed as T scores (M = 50, SD = 10). Self–parent agreement scores are expressed as Pearson r values. ACI = adolescent with cochlear implant; AH = adolescent with hearing.

^aMean values for self-reported and parent-reported anxiety levels were significantly different from one another.

^*p < .05 based on one-tailed tests;

^**p ≤ .01.

Figure 2.

Behavior Assessment System for Children–Second Edition (BASC-2) agreement between self- and parent-reporting of adjustment. In the AH sample alone, significant differences were uncovered between self- and parent-reported scores on the Anxiety and Hyperactivity subscales. ACI = adolescent with cochlear implant; AH = adolescent with hearing. *p < .05; **p ≤ .01.

Next, the magnitude of self–parent agreement was evaluated. In the ACI sample, correlations between self- and parent-reported adjustment scores were statistically significant for the Attention Problems, Depression, and Hyperactivity subscales, suggesting significant agreement between reporters. In the AH sample, correlations between self- and parent-reported scores were statistically significant for the Attention Problems, Atypicality, and Somatization subscales, also suggesting agreement between reporters. Notably, a significant correlation between self- and parent-reported scores in the AH sample was also uncovered for the Anxiety subscale, but the actual mean subscale scores were significantly different across self- and parent-reporting, suggesting divergence across reporters.

Discussion

The present study examined three primary questions: (a) Do samples of ACI and AH self-report differences in psychosocial adjustment? (b) Are there differential associations between self-reported psychosocial adjustment and behavioral measures of neurocognition for ACI versus AH samples? (c) What is the degree of agreement between self- and proxy-parent–reporting of psychosocial adjustment?

Self-Reported Adjustment

Regarding our first research question, our findings indicate that ACIs, when compared to our sample of AH, self-reported more difficulties in several areas of psychosocial adjustment, including atypical behaviors, depression, interpersonal relations, and self-reliance. Moreover, based on self-reports, ACIs were more likely than their AH peers to display at-risk to clinically significant delays in atypicality, social stress, interpersonal relations, and self-reliance. However, it should be noted that the ACI sample self-reported mean adjustment scores within the normative mean for all subscales on the BASC-2. Overall, this pattern of results suggests that approximately 75% of the ACIs in our sample self-report comparable adjustment when compared to their AH peers and a normative sample of adolescents. Notably, for approximately a quarter of ACIs, their self-reports place them at greater risk than their AH peers for poorer psychosocial outcomes centering on difficulties with self-reliance, forming and maintaining healthy social relationships, managing stress, and exhibiting typical behavioral patterns.

Group differences were observed on the Self-Reliance subscale of the BASC-2 such that the ACI sample self-reported poorer adjustment as compared to their AH peers. This subscale measures adolescents' self-confidence and assurance in their ability to make decisions. Our findings are consistent with prior literature indicating that ACIs may struggle with self-consciousness and worries about interpersonal relations (Punch & Hyde, 2005, 2011). Adolescents who self-report lower levels of self-reliance tend to also self-report higher levels of depression and have higher levels of attention problems and hyperactivity based on parent-reports (Kamphaus et al., 2003), and data from our ACI sample supports this finding. That is, our sample of ACIs self-reported more difficulty in the domains of self-reliance and depression than their AH peers, and parents reported that ACIs had more attention problems and hyperactivity than their AH peers.

Group differences were also observed on the Atypicality subscale of the BASC-2 such that the ACI sample self-reported poorer adjustment as compared to their AH peers. This subscale measures the tendency of a child to behave in ways that deviate significantly from the norm and are considered odd or unusual in nature. Differences between ACI and AH samples on the Atypicality subscale may reflect the sensory sequelae of hearing loss, as opposed to atypical behavior as a result of emotional or psychological problems. Because several items on the Atypicality subscale may reflect sensory and related consequences of hearing loss specifically, this measure may not be a valid index of psychosocial functioning for ACIs because of the confound with hearing loss and the poor resolution of the CI signal. Alternatively, the Atypicality subscale may identify some behaviors that are consequences of hearing loss, but appear atypical to hearing peers, potentially increasing challenges with social stresses and interpersonal relations (which are found to be at-risk in the present study). For example, four items on the parent-completed form may reflect communication, social, or comprehension problems arising from hearing loss: appearing confused (which may reflect communication challenges or missing auditory stimuli/information), babbling to self (which may reflect subvocal rehearsal as a strategy during working memory tasks), seeming unaware of others (which may result from missing auditory cues), and having disorganized speech (which may result from speech production or language comprehension challenges). The self-report BASC-2 contains one item (“people think I'm strange”) that may reflect poor speech perception and disturbances in speech intelligibility, instead of atypicality based on behavioral problems or social awkwardness. Future research should examine individual items, and compute subscale scores after removing items that may be confounded with hearing ability and speech perception.

We then examined if sex differences existed in the manifestations of poorer psychosocial adjustment in ACIs and found that female ACIs were significantly more likely to self-report poorer relationships with their parents and personal adjustment than their female AH peers, and poorer self-esteem than their male ACI peers. These findings are important because female ACIs with lower self-esteem and poorer relationships with their parents may struggle with social interactions with peers, leading to social isolation and/or difficulties in developing supportive friendships. Of the few studies that report sex differences, Huber (2005) also found that female ACIs self-reported poorer self-esteem as compared to their female AH peers. Large-scale studies with AH samples consistently indicate that female AH self-report poorer self-esteem than their male counterparts (Bleidorn et al., 2016; Kling et al., 1999). We speculate that these sex differences likely reflect actual behavioral differences rather than differences in self-reporting. This is based on the finding that the female ACI sample self-reported poorer adjustment compared to both their female AH peers and their male ACI peers. This pattern suggests that the sex differences are not solely due to self-awareness or willingness to admit difficulties, but may indeed indicate behavioral differences. Reasons for these sex differences are considerable and range from potential appearance expectations, biological and hormonal changes as a result of puberty, changes in the school environment (transition from middle school to high school), or better social awareness (male adolescents may be more likely to cope by engaging in distancing behaviors such as using humor or not thinking about social ramifications “too much”; see Lazarus & Folkman, 1984, for more information about the adaptive benefits of denial).

Associations Between Self-Reported Adjustment and Neurocognition

Analyses investigating the second research question (relations between neurocognitive and self-reported psychosocial outcomes) demonstrated that, even after controlling for nonverbal intelligence, self-reported psychosocial adjustment was related to language and EF. However, the pattern of these relations was different for ACIs and their AH peers. In ACIs, better language and verbal working memory skills were associated with better self-reported psychosocial adjustment on multiple, broad BASC-2 indices, including internalizing problems, emotional symptoms, and personal adjustment. Furthermore, ACIs with better inhibition–concentration skills self-reported better psychosocial adjustment on all BASC-2 indices, including fewer school problems, fewer internalizing problems, fewer inattention/hyperactivity, fewer emotional symptoms, and stronger personal adjustment. On the other hand, relations between neurocognitive measures and psychosocial adjustment were, for the most part, absent in AH peers, with the exception of fewer self-reported school problems being associated with better language, verbal working memory, and inhibition–concentration skills. Thus, the pattern of results suggests that neurocognitive functions relate predominantly to academic achievement in the AH sample but are more broadly related to emotional–behavioral–social adjustment in the ACI sample.

Associations between EF skills and clinical manifestations of social, emotional, and behavioral problems in at-risk samples are well documented (see Snyder et al., 2015, for a review). Therefore, it was not surprising that ACIs who displayed executive dysfunctions were more likely to self-report broadly poorer psychosocial adjustment. On the other hand, EF skills are particularly crucial for learning and school performance (Barkley, 2006), and language is a core component of most academic subjects. Thus, it is not surprising that EF and language were related to ratings of school problems in the AH sample. However, the failure to find relationships between EF and language skills and other domains of psychosocial functioning in the AH sample suggests that language and EF skills may contribute to psychosocial functioning differentially in ACIs and AH peers. Specifically, it appears that proficiency in language and EF skills is particularly important for global psychosocial adjustment in ACIs, perhaps because ACI users access neurocognitive abilities when engaging in coping strategies to address challenges with social relationships, communication, language/learning, and self-regulatory behavior. For AH peers who have highly proficient language and EF skills, coping with everyday challenges may be less effortful and more automatic, requiring fewer processing resources. Studies are currently underway in our lab to identify how psychosocial adjustment may develop through multiple pathways involving language and EF skills in samples of ACI and AH peers.

Agreement Between Self- and Parent-Reporting

The final research aim of this study was to examine the degree of agreement between self- and parent-reported psychosocial adjustment in the ACI sample, since self- and parent-ratings of social and emotional adjustment can vary as a function of awareness, expectations, and other factors. In the ACI sample, findings revealed no statistically significant disagreements between mean levels of self- and parent-reported psychosocial adjustment. It is possible that because of delays with EF, CI users may require more supportive parental monitoring and involvement in daily life and, as a result, parents may be more aware of their adolescent's experience and perspectives leading to more agreements when reporting behavior. Specifically, data from the ACI sample indicates moderate to strong agreement between self- and parent-reported attention problems, depression, and hyperactivity. It is possible that parents of ACIs are more aware of and sensitive to symptoms of attention problems and hyperactivity because these are areas where CI users are at a greater of risk of developing poorer outcomes following implantation (Castellanos et al., 2018), and/or because these symptoms are more behaviorally evident than other domains of internal emotional adjustment and provide opportunities for parents to shape standards of behavior. The remaining three subscales, which assessed symptoms of anxiety, atypicality, and somatization, showed weak agreement between self- and parent-reporting, with no statistically significant discrepancies observed between reporters. These correlational findings are consistent with previously reported levels of agreement between self- and parent-reports (Renk & Phares, 2004), and may reflect different thresholds for reporting symptoms of anxiety between adolescents and parents and/or parents' differential awareness of adolescents' internal emotional states and behaviors outside of the home (Bird et al., 1992). It is also possible that parents may interpret their adolescent's behavior through the lens of their own concerns, further contributing to the discrepancies observed within these scales. For example, if a parent feels anxious about how their ACI may interact with peers in social gatherings, they may endorse items about anxiety as more severe, even if the ACI does not exhibit significant anxiety. Overall, these factors underscore the importance of considering multiple perspectives when assessing psychosocial adjustment in ACI samples.

In AH samples, extensive prior research suggests that agreement between adolescents' self- and parent-reported social behaviors falls within the weak to moderate range (Achenbach et al., 1987; Gresham et al., 2010; Miller et al., 2014). For instance, a meta-analysis of 74 studies examining self- and parent-reported social skills revealed weak agreement between self- and parent-reports, with average correlations of r = .20 (Renk & Phares, 2004). In the present study, agreement between levels of self- and parent-reported psychosocial adjustment in the AH sample also fell within the weak to moderate range, replicating previously published findings. However, a notable distinction emerged across our samples of adolescents: significant disagreement between mean levels of self- and parent-reported psychosocial adjustment were observed in our AH sample, but not in our ACI sample. These results are consistent with findings reported by Huber, Burger, et al. (2015) in which self-parent agreement was significantly lower in AH peers than in their ACI sample. Within our sample of AH, disagreement in self- and parent-reporting centered on levels of anxiety and hyperactivity, and in both of these cases, parents tended to provide scores in the better-adjusted direction relative to the self-reports. Findings from our AH sample are consistent with Upton et al.'s (2008) meta-analysis of health-related quality of life outcomes indicating that parents of nonclinical samples tend to endorse items in the better-adjusted direction as compared to self-reports.

Study limitations should be considered when interpreting and generalizing the present findings. First, we included children implanted up to age 6.5 years to be more inclusive of children who are late identified. Second, our correlational analyses between self-reported psychosocial adjustment and neurocognitive functioning do not reveal the directional nature of associations, which are likely bidirectional. Studies are currently ongoing in our laboratory to longitudinally examine how growth in neurocognitive functioning may underlie differences in psychosocial adjustment; for example, how proficient EF skills may serve as a protective factor for reducing symptoms of hyperactivity. Third, we focused on six clinical domains of psychosocial adjustment when examining agreement between self- and parent-reporting. Future studies should examine a broader set of psychosocial domains. Finally, our quantitative approach to assessing psychosocial adjustment may benefit from qualitative semistructured interviewing of the parent–adolescent dyad.

In summary, ACIs report comparable psychosocial adjustment to AH peers in many areas, but self-reliance, communication, and social functioning may be at risk for suboptimal outcomes. Better language and EF skills were protective for broad domains of psychosocial adjustment in ACIs but were only protective of school problems in AH peers. Interventions aimed at improving psychosocial adjustment in ACIs could focus on enhancing personal empowerment, resilience, and fostering positive social relationships. This could include social skills training, group therapy sessions, and activities that promote teamwork communication and problem-solving (EF) skills. ACIs and their parents showed moderate to strong agreement in ratings of most domains of psychosocial adjustment. In contrast, self-parent agreement diverged more in the AH sample, particularly in the areas of anxiety and hyperactivity. Importantly, parent-completed behavioral checklists were found to be effective for obtaining reliable information concerning observable (strong self-parent agreement on the Attention Problems and Hyperactivity subscales) and less observable (moderate self-parent agreement on the Depression subscale) psychosocial adjustment in our sample of ACIs. These results support the potential clinical utility of using ratings from multiple informants in order to gain a more comprehensive understanding of psychosocial functioning in ACIs. They also highlight the potential for identifying and improving adjustment in at-risk adolescents by improving language and EF skills.

Data Availability Statement

Data are available to individuals within the scientific community upon request from the corresponding author.

Funding Statement

This research was supported by grants from the National Institute on Deafness and Other Communication Disorders: T32 DC00012 awarded to D.B.P.; R01 DC015257 awarded to D.B.P. and W.G.K.; and R21 DC016134 awarded to I.C., and the National Center for Advancing Translational Sciences: TL1 TR001107 supported I.C.