Abstract Speech intelligibility, or how well a speaker’s words are understood by others, affects listeners’ judgments of the speaker’s competence and personality. Deaf cochlear implant (CI) users vary widely in speech intelligibility, and their speech may have a noticeable “deaf” quality, both of which could evoke negative stereotypes or judgments from peers. In this study, college students with typical hearing (TH) used semantic differential scales to rate speech samples of highly-intelligible TH young adults and age-matched CI users with high or low intelligibility (CI-Hi, CI-Lo) on personality traits related to competence (intelligence, achievement), friendship skills (friendliness, popularity), and attractiveness as a friend (extraversion, dependability). Judges rated TH positively, CI-Lo negatively, and CI-Hi as intermediate, even though CI-Hi were as intelligible as TH. Both CI user groups were rated as friendly but unattractive as friends (insecure, shy, boring, unpopular, does not “sound like someone who could be my friend”), underlining the role of deaf speech quality in peer judgments. Such negative first impressions are likely to affect CI users’ social interactions and friendships, highlighting the importance of speech intelligibility and quality for CI users and calling for education on deafness and deaf speech for TH peers. Good speech intelligibility—speech that is easy for others to understand—is an important speech-language outcome for deaf people with cochlear implants (CIs), who vary widely in cognitive, linguistic, and social skills (for overviews, see Dammeyer, 2010; Pisoni et al., 2008). Good speech intelligibility entails mastery of multiple foundational spoken language skills, including speech perception, phonological knowledge, and motor control of articulation, timing, and intonation (Chin, Bergeson, & Phan, 2012; Ertmer, 2011; Monsen, 1978; Montag, AuBuchon, Pisoni, & Kronenberger, 2014), and it is positively related to a range of cognitive skills (e.g., visual attention and behavioral inhibition, Horn, Davis, Pisoni, & Miyamoto, 2004) and psychosocial behaviors (e.g., socialization and emotional adjustment, Freeman, Pisoni, Kronenberger, & Castellanos, 2017). Speech intelligibility is also related to social and emotional development: among children and teens with hearing loss, speech intelligibility has been linked to social integration with hearing peers (Markides, 1989), self-image, social, psychosocial, and emotional adjustment (Freeman et al., 2017; Polat, 2003), sense of coherence, and loneliness (Most, 2007). Feedback and judgments from interlocutors may underlie these links, for example through misunderstandings, requests for repetition, comments on linguistic skills, and judgments about the CI users’ personality, cognitive abilities, or social position. Such judgments—particularly from peers—can affect CI users’ self-concept as well as the quality and quantity of their social interactions and friendship ties (Batten, Oakes, & Alexander, 2014; Markides, 1989; Most, 2010). The present study examined the first impressions that peers make upon hearing speech from CI users compared to speakers with typical hearing. These first impressions included judgments of competence (intelligence, achievement), personality traits that are desirable in friends (friendliness, extraversion, dependability), and whether the listener would consider the speaker as a potential friend—judgments that are likely to affect subsequent friendship-formation in natural social interaction. Because the focus of this study was on judgments that are likely to affect friendship with peers, both the speakers and judges in this study were in the same age range—early (or “emerging”) young adulthood, a period in which friendship quality has a crucial influence on social adjustment and transition to independence from parents. Methods were adapted from three prior studies that compared adults’ or teens’ impressions of children with and without hearing loss. Related Work Past studies have found links between speaker hearing status, speech intelligibility, and listener-perceived competence and personality traits. Three such studies provided the primary theoretical and methodological background for the present study. In all three, TH adults or teens listened to speech samples from children with or without hearing loss and rated them on various personality traits using semantic differential scales, which display a scale of points (typically 4–7) arranged linearly between two opposing anchor traits (e.g. “friendly…unfriendly”). The present study is the first to apply these methods to young adult speakers; it also used young adult listeners and adaptations of the scales used in the three previous studies. In Blood, Blood, and Danhauer (1978), college students rated the personality (kindness, friendliness, extraversion, etc.), intelligence, achievement, and physical attractiveness of deaf, hard-of-hearing (HH), and typically hearing (TH) elementary school boys after listening to audio recordings of their speech and viewing photos of them with or without a hearing aid. Deaf children were rated the most negatively in all four areas, TH the most positively, and HH as intermediate. Note that whereas the paper did not report hearing aid use and speech intelligibility measurements, it is likely that the deaf group had the poorest speech intelligibility (Smith, 1975), especially given that CIs were not yet available for children (Clark, 1997). Within each hearing group, children were rated more negatively when pictured wearing a hearing aid, which the authors interpreted as a sign of likely discrimination against hearing aid users. To avoid this priming bias, judges in the present study were not told about the hearing status of the speakers prior to the personality rating task. In Most, Weisel, and Lev-Matezky (1996), TH young adults rated the speech intelligibility, cognitive competence (intelligence, achievement), and personality (extraversion, self-confidence, sociability, etc.) of deaf, HH, and TH elementary/middle-school children after listening to audio recordings of their speech. Judges who had no experience with deaf/HH speech rated deaf children the most negatively in all three areas, TH children the most positively, and HH children as intermediate. This explicit link between deaf/HH speech intelligibility and personality/competence ratings formed the basis for the predictions in the present study. Most, Weisel, and Tur-Kaspa (1999) used similar procedures with high school students as judges. The speakers were one TH boy who was rated in the 1996 study as highly intelligible, one HH boy rated as moderately intelligible, and one deaf boy rated as poorly intelligible. They substituted the cognitive competence scales from the 1996 study for an emotional-behavioral composite with scales focusing on judges’ emotional reactions to interactions with the speaker (nervousness, fondness) and their interest in a potential relationship (friend, classmate, date). As with the adult judges in the 1996 study, teens without deaf/HH classmates rated the TH child highest in all three areas, the deaf child the lowest, and the HH child as intermediate. Similar to the Most et al. (1999) design, the present study divided speakers by speech intelligibility as determined in a prior study, and the rating task included traits related to friendship. However, the present study calculated speech intelligibility as the proportion of speech that multiple TH transcribers correctly recognized rather than using a single holistic rating, and trait ratings from individual scales were not combined into composite scores. Both Most et al. studies also included judges who had substantial experience with deaf/HH speakers. Teachers and clinicians in the 1996 adult study rated the competence and personality of deaf children as highly as HH children, indicating that deaf contact improved adults’ judgments of deaf children (but not to the level of TH children). Similarly, students with at least two deaf/HH classmates in the 1999 teen study rated the personality traits of the HH child as low as the deaf child, but their ratings on the emotional-behavioral composite did not differ from students without deaf classmates, suggesting that contact with deaf/HH peers did not affect teens’ willingness to make friends with unfamiliar deaf/HH speakers. However, in keeping with the present study’s focus on first impressions of unfamiliar TH peers, the majority of whom do not have much prior contact with young deaf people, judges with extensive experience with deaf/HH speech were not included. In contrast to all three prior studies, speakers and judges in the present study were in the same young-adult age range, and the two speaker groups with hearing loss were both deaf CI users who varied in speech intelligibility, rather than varying in both severity of hearing loss and intelligibility. (Unaided, their hearing loss was at least 85 dB, similar to the deaf groups in past studies.) Young adulthood is characterized by increasing independence from family, when peer friendships take on greater importance (Shulman, 1975). In Erikson’s (1950) psychosocial development stage framework, a primary concern during young adulthood is the formation of intimate friendships and romantic relationships. More recently, Arnett (2004) defined the late teens through mid-20s as emerging adulthood, a developmental stage between adolescence and adulthood involving a transition to independence from parents and exploration of identity, relationships, and career options. During this period, most Americans today attend college (Census Bureau, 2015), where hearing students may encounter CI users and deaf speech for the first time. CI users are also attending college at similar rates (Spencer, Tomblin, & Gantz, 2012). Because new peer friendships in the first year of college influence adjustment to academic life (Buote et al., 2007; Swenson, Nordstrom, & Hiester, 2008), the first impressions and judgments that facilitate or hinder acceptance and friendship with new, predominantly hearing, classmates may be particularly important for CI user college students, in both their social adjustment and transition to full adulthood. In the present study, all judges were college students in their late teens or early 20s who responded to speech samples from CI users in the same young/emerging adult age range. The judges’ instructions identified the speakers as “about their age” but not as college students in order to avoid biasing their impressions of the speakers’ intelligence. Research Questions and Predictions The present study addressed the following research questions. Predictions for the first two questions paralleled the global results of the above studies; the second two questions explored more local components of peer-rating behavior. Research Question 1: How are voice-based personality trait peer-ratings affected by speakers’ hearing status? Following the above studies, judges were predicted to rate TH speakers more positively than CI users. Research Question 2: How are voice-based personality trait peer-ratings affected by speech intelligibility? Following the above studies, a positive relationship was predicted such that ratings for highly-intelligible CI users (CI-Hi) would be more positive than for less-intelligible CI users (CI-Lo). If the prediction for Research Question 1 also found support, the three speaker groups would show separation in ratings, with TH rated most positively, CI-Lo most negatively, and CI-Hi intermediate. However, if speech intelligibility influenced ratings more than hearing status, the two highly-intelligible groups, CI-Hi and TH, would pattern together as more positively rated than CI-Lo. Research Question 3: Among CI users, are some personality traits rated more negatively than others? For example, CI users might be rated low on intelligence but high on friendliness, possibly reflecting an association of deaf speech with stereotypes of people with intellectual disabilities. Research Question 4: Are some rating choices made more confidently than others? Response latency, the time between the display of a rating scale and the judge’s selection, can serve as a measure of the strength or certainty of the judge’s impression, or the ease or automaticity of associating the speaker with the rating (Draisma & Dijkstra, 2004; Mulligan, Grant, Mockabee, & Monson, 2003). Thus, shorter latencies would indicate easier decisions and choices with closer fits to the judge’s overall impression of the speaker. For example, if a speaker evoked strong impressions of low intelligence but no particular impression of supportiveness, judges would quickly rate the speaker as “dumb” but take more time in choosing a rating for supportiveness. Method The present study tested the above predictions using a within-subjects correlational design. One group of 40 college student participants rated 12 young-adult speakers on 18 personality traits using five-point semantic differential scales. The speakers varied in hearing status (CI, TH) and speech intelligibility (High, Low) to form three groups of four speakers each: TH (highly intelligible), CI-Hi, CI-Lo. The personality scales represented five categories of traits (Competence, Extraversion, Dependability, Friendship, and Voice) and were adapted from items in Blood et al. (1978), Most et al. (1996, 1999), and previous studies on friendship expectations that identified traits possessed by good friends or considered desirable in ideal friends (see review in Hall, 2011). The analyses examined relationships between group patterns of personality trait ratings and speakers’ hearing status and/or speech intelligibility. Participants (Judges) Participant judges were 40 college students (23 female, 17 male) who received credit in introductory psychology courses and who self-reported as monolingual native English-speakers age 18–23 (M = 19 years) with no known hearing, speech, or cognitive disorders. Judges were primarily of European descent, from the American Midwest, and with little contact with people who are deaf or hard-of-hearing. (One additional male judge was excluded due to excessive time spent completing the task, suggesting inattention or cognitive difficulty.) Materials Stimuli and speaker groups The speech samples to be rated by the judges originated from audio recordings of 12 young adults who each completed a sentence-repetition task called the McGarr Sentence Intelligibility test (McGarr, 1981) as part of a project on long-term outcomes for CI users (see Kronenberger, Pisoni, Henning, & Colson, 2013; Montag et al., 2014). The McGarr test consists of 36 simple sentences which vary in length and contextual predictability; each sentence is read aloud from a printed card by the test examiner and then repeated aloud by the test participant. McGarr materials have been used in previous work to assess the speech intelligibility of deaf speakers with CIs compared to speakers with typical hearing by replaying the speakers’ sentences to naïve listeners, whose transcriptions were scored as to the percentage of words correctly recognized in each sentence (Freeman et al., 2017; Geers, 2002; Montag et al., 2014; Tobey, Geers, Sundarrajan, & Shin, 2011). That is, speakers were considered intelligible to the degree that unfamiliar listeners correctly recognized and transcribed their words. The current study calculated two speech intelligibility scores for each speaker: McGarr intelligibility and perfect-sentence intelligibility. McGarr intelligibility reflects the conventional scoring method, the percentage of a speaker’s words recognized over all 36 sentences, averaged across three transcribers with typical hearing and no experience with deaf/hard-of-hearing speech. Perfect-sentence intelligibility is a stricter measure, the percentage of a speaker’s 36 sentences that all three listeners transcribed perfectly (Freeman et al., 2017). (Note that in the current study, only one listener transcribed speakers with typical hearing.) From a pool of 33 young adults (age 18–27) with McGarr recordings, two male and two female speakers were selected to represent each of three groups: typical-hearing (TH), highly-intelligible CI users (CI-Hi), and less-intelligible CI users (CI-Lo). Table 1 lists the intelligibility scores (mean and range) of each speaker group. Note that McGarr intelligibility was comparable between TH and CI-Hi, but CI-Hi perfect-sentence intelligibility was lower (although still good), while CI-Lo was much lower and more variable on both measures. A comparison of the intelligibility scores of all sentences pooled for all speakers in each group supported these group relationships: CI-Lo sentences were less intelligible than CI-Hi and TH sentences (both p < .001, Wilcoxon rank-sum tests), but CI-Hi sentences were only slightly less intelligible than TH (p < .05). Thus, TH and CI-Hi differed primarily in hearing status, CI-Hi and CI-Lo differed in intelligibility, and TH and CI-Lo differed on both dimensions. (There were no TH speakers with low intelligibility.) Table 1. Intelligibility scores by speaker group (mean, range) Speaker group McGarr intelligibility (%) Perfect-sentence intelligibility (%) TH 96 (93–98%) 85 (78–92%) CI-Hi 95 (93–97%) 70 (61–78%) CI-Lo 71 (58–82%) 23 (6–31%) Speaker group McGarr intelligibility (%) Perfect-sentence intelligibility (%) TH 96 (93–98%) 85 (78–92%) CI-Hi 95 (93–97%) 70 (61–78%) CI-Lo 71 (58–82%) 23 (6–31%) Note: Values are mean (range); McGarr = % of words recognized averaged across transcribers; perfect-sentence = % of sentences perfectly recognized by all transcribers. Rating scales Judges rated speakers’ personalities using five-point semantic differential scales from five categories. Table 2 lists the 18 personality traits by category, together with the on-screen prompts and scale anchors that judges viewed (see Figure 1 for an example scale). Table 2. Personality traits presented in the rating task Trait Prompt Scale anchors Competence smart This person sounds… dumb…smart wise This person sounds like someone who makes… wise decisions…stupid decisions achiever This person sounds like a… high achiever…low achiever hard-working This person sounds… lazy…hard-working Extraversion confident This person sounds… insecure…confident outgoing This person sounds… outgoing…shy cheerful This person sounds… cheerful…gloomy fun This person sounds… boring…fun Dependability mature This person sounds… mature…immature stable This person sounds… unpredictable…emotionally stable reliable This person sounds… reliable…unreliable supportive This person sounds… self-centered…supportive Friendship friendly This person sounds… friendly…unfriendly sociable This person sounds… stand-offish…sociable popular This person sounds like someone with… many friends…few friends my-friend This person sounds like someone I could be friends with. disagree…agree Voice clear-speech This person’s speech is… easy to understand…hard to understand normal-voice This person’s voice sounds… odd…normal Trait Prompt Scale anchors Competence smart This person sounds… dumb…smart wise This person sounds like someone who makes… wise decisions…stupid decisions achiever This person sounds like a… high achiever…low achiever hard-working This person sounds… lazy…hard-working Extraversion confident This person sounds… insecure…confident outgoing This person sounds… outgoing…shy cheerful This person sounds… cheerful…gloomy fun This person sounds… boring…fun Dependability mature This person sounds… mature…immature stable This person sounds… unpredictable…emotionally stable reliable This person sounds… reliable…unreliable supportive This person sounds… self-centered…supportive Friendship friendly This person sounds… friendly…unfriendly sociable This person sounds… stand-offish…sociable popular This person sounds like someone with… many friends…few friends my-friend This person sounds like someone I could be friends with. disagree…agree Voice clear-speech This person’s speech is… easy to understand…hard to understand normal-voice This person’s voice sounds… odd…normal Figure 1. View largeDownload slide Rating task instructions, on-screen (top) and oral supplement script (bottom). Figure 1. View largeDownload slide Rating task instructions, on-screen (top) and oral supplement script (bottom). Categories of personality traits grouped related constructs; each of the four personality categories included two concepts with two trait scales each. The Competence category included intelligence and work ethic/achievement, traits positively related to social popularity among teens (Berndt & Das, 1987; Meijs, Cillessen, Scholte, Segers, & Spijkerman, 2010) and negatively to perceived disability, including deafness (Berkay, Gardner, & Smith, 1995). (Similar traits appeared in the cognitive competence composite in Most et al. (1996) and the intelligence and achievement composites in Blood et al. (1978); many of the traits in the remaining categories appeared in these past studies’ personality composites and the cognitive composite in Most et al., 1999). Extraversion included self-confidence and optimism, traits often possessed by likeable/popular people or considered desirable in friends (Sprecher & Regan, 2002; van der Linden, Scholte, Cillessen, te Nijenhuis, & Segers, 2010). Dependability included emotional stability and supportiveness, traits often possessed by “good friends” or considered desirable in ideal friends (Buhrmester, Furman, Wittenberg, & Reis, 1988; van der Linden, et al., 2010; see also studies in Hall, 2011). Friendship included overt assessments of friendliness, or the speaker’s desire and attempt to make friends, and popularity, the speaker’s success of making friends or gaining the acceptance of others, including the judge personally (similar to the emotional-behavioral scales in Most et al., 1999). The fifth category contained two Voice items designed to reflect speech intelligibility (clear/understandable-speech) and an impression of deaf speech characteristics (normal/odd voice). Deafness was not explicitly mentioned during the rating task to avoid biasing judges or activating non-speech-related stereotypes of deaf people. Procedure Indiana University’s institutional review board approved all study protocols, and participants gave informed consent at the beginning of the rating sessions, which the author conducted in a speech lab at the beginning of the school year. Rating sessions included 1–4 judges at a time and lasted about 30 min. Each judge sat at a computer with high-quality circumaural headphones set to a comfortable loudness of approximately 65 dB. The following instructions appeared on each computer screen and were reviewed orally by the experimenter, supplemented by the additional oral instructions below. The practice round presented six sentences produced by one male TH speaker; the recordings originated from an unrelated sentence-reading task, and neither the sentences nor the speaker appeared in subsequent test materials. In the test rounds, the presentation of speakers differed for each judge and was pseudo-randomized so that no two speakers from the same group appeared sequentially. For each speaker, six sentences of varying lengths and contextual predictability were presented in random order, separated by 500 ms. Judges heard each sentence once across the first six speakers and once across the last six speakers. After hearing a speaker’s six sentences, judges rated the speaker on each personality trait listed in Table 2. Half the traits always had the positive anchor displayed on the right end of the scale, half on the left. The first 15 trait scales in Table 2 appeared in random order, followed by the last three in fixed order (my-friend, clear-speech, normal-voice), and then a prompt for judge-supplied descriptions or comments (“Can you think of anything else to describe this person or voice?”). Finally, judges completed a short questionnaire on paper that asked about the difficulty of the task, strategies for choosing ratings, experience with people who are deaf or hard-of-hearing, and demographics. Data analysis Each rating response received a numeric score from 1 to 5, with 1 corresponding to the most negative option and 5 to the most positive (the trait names in Table 2 identify the positive scale anchors). Ratings were averaged across judges for each scale and across all personality traits (i.e., excluding the two Voice traits) for each speaker group. To obtain a measure of rating strength or confidence, response latencies were recorded as the time (ms) between the appearance of a rating scale on the computer screen and the corresponding rating selection. Latencies were averaged across judges within each speaker group and each rating level (1–5). Because ratings and response latencies were not normally distributed, analyses used non-parametric statistical tests. Wilcoxon two-tailed rank-sum tests compared judge and speaker groups on personality ratings and response latencies. Spearman rank correlations tested the relations between speech intelligibility and personality ratings. Results Ratings Figure 2 plots the mean ratings for each speaker group on each trait, averaged across all 40 judges, with more positive ratings higher on the y-axis. On most traits, the TH speaker group was rated most positively and the less-intelligible CI speaker group (CI-Lo) most negatively. The highly-intelligible CI speaker group (CI-Hi) was rated between the other two groups on most traits but similar to CI-Lo as low on extraversion traits. Figure 2. View largeDownload slide Mean trait ratings by talker group. Figure 2. View largeDownload slide Mean trait ratings by talker group. The TH speaker group was rated more positively than each CI-user group on each trait (all p < .05) except supportive/self-centered. The ratings for supportiveness did not differ between the three speaker groups and were near the middle of the scale, suggesting that the speech samples did not evoke clearly positive or negative impressions of this trait. The two CI groups did not differ on three additional traits (hard-working, friendly, sociable). CI-Lo was rated slightly but significantly more positively than CI-Hi on two extraversion traits (cheerful, fun, both p < .05). However, the highly-intelligible CI speaker group (CI-Hi) was rated significantly more positively than the less-intelligible CI speaker group (CI-Lo) on the majority of traits (all p < .05). Speaker groups were more easily distinguished on some traits than others. As seen in Figure 2, TH and CI-Lo were clearly separated on the voice traits (normal/odd voice and clear/unclear speech), as expected, and on most other traits, except supportive and friendly. As mentioned above, supportiveness may not have been determinable for any speaker, but the relatively high rating for friendliness was a strong departure from the CI-Lo group’s low ratings on other traits. Ratings for the CI-Hi speaker group were intermediate but somewhat closer to ratings for TH than CI-Lo on most competence, dependability, and voice traits, but CI-Hi patterned more closely with CI-Lo on extraversion and friendship traits. In short, judges rated TH speakers positively, CI-Lo as friendly but otherwise lacking in desirable traits, and CI-Hi as moderately competent, dependable, and clear, but as friendly yet unpopular and introverted as CI-Lo. Note that when TH speakers were one of the first three presented, their ratings were more negative than for subsequent TH speakers, and when CI-Hi speakers were one of the first two presented, their ratings were more negative than for subsequent CI-Hi speakers (both p < .01), but speaker order did not similarly affect CI-Lo speakers. Also note that ratings for the two TH male speakers were more negative than for the two TH female speakers on all traits (all p < .05). Ratings for the TH males were as low as for the CI-Hi group on nine scales (wise, achiever, hard-working, cheerful, fun, reliable, supportive, friendly, sociable). Ratings for the CI speaker groups did not differ substantially by speaker gender, and ratings did not differ by judge gender for any speaker group. Intelligibility and Ratings Table 3 lists mean intelligibility scores, ratings on Voice traits (clear-speech, normal-voice), and overall ratings by speaker group. McGarr scores are the mean percentage of a speaker’s words which multiple transcribers recognized in a previous study. Perfect-sentence intelligibility is the proportion of a speaker’s sentences which all listeners transcribed perfectly in the previous study. The clear-speech and normal-voice ratings are mean ratings across all judges in the current study on the questions about understandable speech and normal/odd-sounding voice. The overall mean rating was calculated for each speaker across all traits and judges. All of these measures were highly correlated (all ρ > .75, p < .001), indicating that judges viewed more intelligible speakers as having clear, normal speech and rated them more positively overall. However, while the CI-Hi group was equal to TH in McGarr intelligibility and closer to TH than CI-Lo on perfect-sentence intelligibility, clear-speech, and normal-voice ratings, the overall ratings for CI-Hi were nearly equidistant from TH and CI-Lo, suggesting that a factor other than speech intelligibility contributed to voice-based personality judgments for this CI user group. Table 3. Mean intelligibility, clear-speech, and overall ratings by speaker group Speaker group McGarr intelligibility (%) Perfect-sentence intelligibility (%) Clear-speech rating Normal-voice rating Overall rating TH 96 85 4.7 4.5 3.6 CI-Hi 95 70 3.9 3.4 2.9 CI-Lo 71 23 1.5 1.5 2.3 Speaker group McGarr intelligibility (%) Perfect-sentence intelligibility (%) Clear-speech rating Normal-voice rating Overall rating TH 96 85 4.7 4.5 3.6 CI-Hi 95 70 3.9 3.4 2.9 CI-Lo 71 23 1.5 1.5 2.3 Response Latencies Mean response latencies decreased slightly over the course of the experiment (ρ = −.19, p < .001), plateauing at the seventh speaker presented. Mean response latencies were faster when rating TH speakers than CI speakers (p < .001) but did not differ between the two CI speaker groups. Across traits, response latencies were faster when giving negative ratings (rating of 1) than positive or intermediate ratings (ratings 2–5, p < .001). Within the TH speaker group, judges gave positive ratings (rating 5) fastest, followed by negative and slightly-positive (ratings 1, 4), followed by slightly-negative and neutral (ratings 2, 3, all p < .01). Response latencies when giving negative, slightly-negative, or neutral ratings (ratings of 1, 2, or 3) did not differ across speaker groups. When giving a slightly-positive rating (rating of 4), response latencies were fastest when rating TH speakers, then CI-Hi, then CI-Lo (all p < .01). Similarly when giving the most positive rating (rating of 5), response latencies were fastest when rating TH speakers (both p < .001), but they did not differ between CI speaker groups. In short, judges quickly chose positive ratings for TH speakers and negative ratings for all speaker groups, while selecting intermediate ratings and positive ratings for CI speakers more slowly. Note that judges gave a majority (54%) of their negative ratings (rating 1) to CI-Lo speakers but a majority (69%) of positive ratings (rating 5) to TH speakers, contributing to the pattern that TH speakers were quickly rated positively and CI speakers were quickly rated negatively, with departures from this pattern selected more slowly. Discussion The present results supported the predictions for all four research questions: (1) ratings for TH speakers were more positive than for CI users overall and on nearly every personality trait, (2) ratings for highly-intelligible CI users (CI-Hi) were more positive than for less-intelligible CI users (CI-Lo) overall and on most individual traits, (3) ratings for CI-Hi (but not CI-Lo) were more negative on some categories (extraversion and popularity) than others, and (4) judges gave positive ratings to TH more quickly than to CI users. The divisions between CI-Lo and TH or CI-Hi are in line with past studies linking speech intelligibility to personality judgments of children (Most et al., 1996, 1999). However, the division between TH and CI-Hi suggests that speech intelligibility was not the only factor that negatively affected peers’ judgments of young adult CI user speech—perhaps some aspect of “deaf” speech quality was detectable in the speech samples even though the words were easily recognizable, as suggested by the moderate rating for CI-Hi on the normal/odd-voice item compared to the very high rating for TH. Additionally, judges may have been sensitive to small differences in intelligibility that separated CI-Hi from TH. The rating for CI-Hi on the clear-speech item is particularly telling in this regard: judges rated CI-Hi speakers as harder to understand than TH, closely mirroring the two groups’ perfect-sentence intelligibility scores but not their nearly-equivalent McGarr scores (Table 3). The overall rating pattern for CI-Hi indicates that TH judges thought these speakers sounded moderately smart and dependable but shy and unpopular—good people but not too appealing as personal friends. The pattern of low ratings for CI-Lo on all traits except friendliness suggests that their speech evoked a stereotype associated with intellectual disabilities—unintelligent and friendly but not truly anyone’s friend. However, this group was not rated as extraverted or confident, traits that might fit with a blissfully friendly stereotype. Response latencies can be interpreted as decision times reflecting a degree of difficulty in making a rating choice, but the difficulty may have more than one cause. If latencies reflect judge confidence or certainty in rating choices (as suggested by slower decision times for intermediate ratings), then judges here were more certain of their positive ratings of TH speakers and negative ratings in general but less certain of positive ratings for CI users. Similarly, decision-making may have been delayed when judges were surprised or found a mismatch between their expectations and the stimuli; for example, upon hearing a CI-Lo speaker, judges might have reacted generally with negative impressions but took an extra moment to consider a more positive rating for certain traits. Implications and Clinical Relevance As some of the first work to examine college students’ speech-based judgments of peer CI users, this study shed light on social challenges that CI users may face upon entering college, as many now do. The personality rating patterns found in this study showed biases against CI users based on their speech, particularly for speakers who were not perfectly intelligible, highlighting the importance of speech intelligibility for CI users. Such biases may hinder friendship development with new classmates and therefore adjustment to college life and the transition toward independent adult living. In addition, the intermediate ratings for the highly-intelligible CI speakers showed that good speech intelligibility was not enough to eliminate biases against deaf speech, further increasing the importance of clinical attention to features of “deaf” speech quality. Listener biases could also be addressed directly. Prior work has found that experience with deaf/HH children/teens improved adults’ and teens’ ratings of deaf speech intelligibility, personal qualities, and abilities (Most et al., 1996, 1999). A little education could produce similar results. The judges in the present study had very minimal contact with deafness or hearing loss, and the few encounters they described mainly involved older adults with whom they “talked louder” or deaf acquaintances who wrote or gestured their needs—in other words, they had very limited exposure to deaf/HH speech. In fact, none of the judges recognized any of the speakers as deaf, even though the CI-Lo speakers had several marked features of deaf speech. Some judges commented that some of the speakers sounded like they had a speech impediment or an intellectual disability. This association of deaf speech quality with intellectual disability may have evoked stereotypes that biased the ratings of CI speakers. Experience with deaf speech and education on the capabilities of deaf people may reduce this bias and improve attitudes toward deafness, including personality ratings. Classmates of deaf students would be particularly important targets for both deafness education and increased interaction, for example through group work with teacher-assigned partners. Such experience might be especially important during elementary and high school, when contact among students is more concentrated and sustained than in college or work settings. Short educational programs might include information about deafness/hearing loss (including CIs and hearing aids), practice with methods that facilitate communication with deaf/HH peers, discussion of misconceptions about deafness and deaf people’s capabilities, examples of ordinary deaf people who are successful in a wide variety of jobs and activities, and testimonials from deaf and hearing friends. Limitations and Future Directions This study was designed to capture first impressions evoked upon hearing an unfamiliar voice, but the laboratory setting and prerecorded speech samples may not accurately reflect ordinary situations in which most first impressions are formed. The speech samples consisted of disconnected sentences rather than meaningful, contextualized exchanges, which normally include visual and social information. Real-world first impressions are also not formed via a checklist of externally supplied adjectives, nor do people normally consciously judge 12 new personalities on the same 18 dimensions in one half hour. The influence of speaker comparison may be inherent in the study procedure, as suggested by the fairly negative ratings given to the first few speakers heard, regardless of speaker group (which differed for each judge), followed by separation between the groups for subsequent speakers. In real-world situations, the traits that are salient to forming first impressions may vary for each person encountered, without comparison to others. There are also many factors beyond speech clarity and intelligibility that contribute to personality judgments, as evidenced by the substantially lower ratings for male compared to female TH speakers on nearly every trait except normal-voice and clear-speech. Gender is often a salient social factor, but it did not affect ratings for the CI user groups. Perhaps individual differences among the four particular TH speakers in this study were coincidentally greater than among the CI user groups—future work should explore this with a larger speaker pool—or perhaps the influences of speech quality and intelligibility overshadowed gender effects in the CI user groups. Some judges commented after they completed the study that it was difficult or unfair to judge people based solely on their voices, especially when they were reading random sentences rather than discussing meaningful content or interacting socially. It could be that impressions in everyday first encounters are more nuanced than those found here because they have the benefit of linguistic content to help judge competence, emotional expression to judge extraversion, body language to judge friendliness, and so on, and judges felt they did not have enough information to form an accurate impression. Some judges commented more specifically that they felt uncomfortable labeling people they perceived to have a disability. Responses to follow-up questions from the experimenter suggested that these judges felt that they should not assume that a disability in one area necessarily resulted in deficits in other areas, or simply that it is not socially acceptable to label minority groups. These attitudes reflect positively on the tolerance and social education of college students, at least in terms of their willingness to express negative views of others in a formal educational setting. However, the consistency in mean rating patterns across judges and of CI users in each group suggests some degree of reliability to their reactions and inferences about personality traits, even when based solely on speech quality. It was these types of “snap judgments” that this study was designed to access. With the assumption that initial biases affect subsequent interactions, expectations, and dispositions, future work should examine the nature of these effects on friendships with CI users. As discussed above, future work should investigate the effects of attitudes toward deafness and education about deafness and deaf speech. As a first step in this direction, a follow-up study is underway to assess the contributions of judge personality traits and preexisting attitudes toward deafness to their personality rating patterns. Future studies may investigate the effect of a short educational program on attitudes toward deafness and subsequent personality ratings of deaf speakers. Finally, this study used only young adult speakers and judges, but friendship-formation and acceptance by peers are also important during childhood and especially teenage years, when attitudes and social routines are strengthened. College student judges in particular may be more tolerant of social differences than teens and young adults in other social surroundings. (However, note that judges did not rate CI users as people who “could be their friends.”) Future work will extend the rating procedures to teens and school-age children. Funding National Institutes of Health—National Institute on Deafness and Other Communication Disorders [R01 DC009581]. Conflicts of interest No conflicts of interest are reported. References Arnett, J. J. ( 2004). Emerging adulthood: The winding road from the late teens through the twenties . New York, NY: Oxford University Press. Batten, G., Oakes, P. M., & Alexander, T. ( 2014). Factors associated with social interactions between deaf children and their hearing peers: A systematic literature review. Journal of Deaf Studies and Deaf Education , 19, 285– 302. doi:10.1093/deafed/ent052. Google Scholar CrossRef Search ADS PubMed Berkay, P. J., Gardner, J. E., & Smith, P. L. ( 1995). The development of the opinions about deaf people scale: A scale to measure hearing adults’ beliefs about the capabilities of deaf adults. Educational and Psychological Measurement , 55, 105– 114. doi:10.1177/0013164495055001010. Google Scholar CrossRef Search ADS Berndt, T. J., & Das, R. ( 1987). Effects of popularity and friendship on perceptions of the personality and social behavior of peers. The Journal of Early Adolescence , 7, 429– 439. doi:10.1177/0272431687074005. Google Scholar CrossRef Search ADS Blood, G. W., Blood, I. M., & Danhauer, J. L. ( 1978). Listeners’ impressions of normal-hearing and hearing-impaired children. Journal of Communication Disorders , 11, 513– 518. doi:10.1016/0021-9924(78)90025-4. Google Scholar CrossRef Search ADS PubMed Buhrmester, D., Furman, W., Wittenberg, M. T., & Reis, H. T. ( 1988). Five domains of interpersonal competence in peer relationships. Journal of Personality and Social Psychology , 55, 991– 1008. doi:10.1037/0022-35220.127.116.111. Google Scholar CrossRef Search ADS PubMed Buote, V. M., Pancer, S. M., Pratt, M. W., Adams, G., Birnie-Lefcovitch, S., Polivy, J., & Wintre, M. G. ( 2007). The importance of friends: Friendship and adjustment among 1st-year university students. Journal of Adolescent Research , 22, 665– 689. doi:10.1177/0743558407306344. Google Scholar CrossRef Search ADS Census Bureau ( 2015). Current Population Survey (CPS), October, 1975 through 2014. U.S. Department of Commerce. Retrieved from Digest of Education Statistics, National Center for Education Statistics: https://nces.ed.gov/programs/digest/d15/tables/dt15_302.30.asp. Chin, S. B., Bergeson, T. R., & Phan, J. ( 2012). Speech intelligibility and prosody production in children with cochlear implants. Journal of Communication Disorders , 45, 355– 366. doi:10.1016/j.jcomdis.2012.05.003. Google Scholar CrossRef Search ADS PubMed Clark, G. M. ( 1997). Historical perspectives. In Clark G. M., Cowan R. S. C., & Dowell R. C. (Eds.), Cochlear Implantation for Infants and Children: Advances (pp. 9– 27). San Diego, CA: Singular Publishing Group. Dammeyer, J. ( 2010). Psychosocial development in a Danish population of children with cochlear implants and deaf and hard-of-hearing children. Journal of Deaf Studies and Deaf Education , 15, 50– 58. doi:10.1093/deafed/enp024. Google Scholar CrossRef Search ADS PubMed Draisma, S., & Dijkstra, W. ( 2004). Response latency and (para)linguistic expressions as indicators of response error. In Presser S., Rothgeb J. M., Couper M. P., Lessler J. T., Martin E., Martin J., & Singer E. (Eds.), Methods for testing and evaluating survey questionnaires (pp. 131– 147). Hoboken, NJ: John Wiley & Sons. Google Scholar CrossRef Search ADS Erikson, E. ( 1950). Childhood and society ( 1st ed.). New York, NY: Norton. Ertmer, D. J. ( 2011). Assessing speech intelligibility in children with hearing loss: Toward revitalizing a valuable clinical tool. Language, Speech, and Hearing Services in Schools , 42, 52– 58. doi:10.1044/0161-1461(2010/09-0081). Google Scholar CrossRef Search ADS PubMed Freeman, V., Pisoni, D., Kronenberger, W. G., & Castellanos, I. ( 2017). Speech intelligibility and psychosocial functioning in deaf children and teens with cochlear implants. Journal of Deaf Studies and Deaf Education , 22. doi:10.1093/deafed/enx001. Geers, A. E. ( 2002). Factors affecting the development of speech, language, and literacy in children with early cochlear implantation. Language, Speech, and Hearing Services in Schools , 33, 172– 183. doi:10.1044/0161-1461(2002/015). Google Scholar CrossRef Search ADS PubMed Hall, J. A. ( 2011). Sex differences in friendship expectations: A meta-analysis. Journal of Social and Personal Relationships , 28, 723– 747. doi:10.1177/0265407510386192. Google Scholar CrossRef Search ADS Horn, D. L., Davis, R. A., Pisoni, D. B., & Miyamoto, R. T. ( 2004). Visual attention, behavioral inhibition and speech/language outcomes in deaf children with cochlear implants. In R. T. Miyamoto (Ed.), Proceedings of the VIII International Cochlear Implant Conference (International Congress Series, Vol. 1273, pp. 332–335). Indianapolis, IN: Elsevier. doi:10.1016/j.ics.2004.07.048. Kronenberger, W. G., Pisoni, D. B., Henning, S. C., & Colson, B. G. ( 2013). Executive functioning skills in long-term users of cochlear implants: A case control study. Journal of Pediatric Psychology , 38, 902– 914. doi:10.1093/jpepsy/jst034. Google Scholar CrossRef Search ADS PubMed Markides, A. ( 1989). Integration: The speech intelligibility, friendship and associations of hearing impaired children in secondary schools. Journal of the British Association of Teachers of the Deaf , 13, 63– 72. McGarr, N. S. ( 1981). The effect of context on the intelligibility of hearing and deaf children’s speech. Language and Speech , 24, 255– 264. doi:10.1177/002383098102400305. Google Scholar CrossRef Search ADS PubMed Meijs, N., Cillessen, A. H., Scholte, R. H., Segers, E., & Spijkerman, R. ( 2010). Social intelligence and academic achievement as predictors of adolescent popularity. Journal of Youth and Adolescence , 39, 62– 72. doi:10.1007/s10964-008-9373-9. Google Scholar CrossRef Search ADS PubMed Monsen, R. B. ( 1978). Toward measuring how well hearing-impaired children speak. Journal of Speech, Language, and Hearing Research , 21, 197– 219. doi:10.1044/jshr.2102.197. Google Scholar CrossRef Search ADS Montag, J. L., AuBuchon, A. M., Pisoni, D. B., & Kronenberger, W. G. ( 2014). Speech intelligibility in deaf children after long-term cochlear implant use. Journal of Speech, Language, and Hearing Research , 57, 2332– 2343. doi:10.1044/2014_JSLHR-H-14-0190. Google Scholar CrossRef Search ADS Most, T. ( 2007). Speech intelligibility, loneliness, and sense of coherence among deaf and hard-of-hearing children in individual inclusion and group inclusion. Journal of Deaf Studies and Deaf Education , 12, 495– 503. doi:10.1093/deafed/enm015. Google Scholar CrossRef Search ADS PubMed Most, T. ( 2010). How does speech intelligibility affect self and others’ perceptions of deaf and hard-of-hearing people? In Marschark M., & Spencer P. E. (Eds.), The Oxford handbook of deaf studies, language, and education ( Vol. 2, pp. 251– 263). New York, NY: Oxford University Press. Most, T., Weisel, A., & Lev-Matezky, A. ( 1996). Speech intelligibility and the evaluation of personal qualities by experienced and inexperienced listeners. Volta Review , 98, 181– 190. Most, T., Weisel, A., & Tur-Kaspa, H. ( 1999). Contact with students with hearing impairments and the evaluation of speech intelligibility and personal qualities. Journal of Special Education , 33, 103– 111. doi:10.1177/002246699903300204. Google Scholar CrossRef Search ADS Mulligan, K., Grant, J. T., Mockabee, S. T., & Monson, J. Q. ( 2003). Response latency methodology for survey research: Measurement and modeling strategies. Political Analysis , 11, 289– 301. doi:10.1093/pan/mpg004. Google Scholar CrossRef Search ADS Pisoni, D. B., Conway, C. M., Kronenberger, W. G., Horn, D. L., Karpicke, J., Henning, S. C., & Hauser, P. C. ( 2008). Efficacy and effectiveness of cochlear implants in deaf children. Deaf Cognition: Foundations and Outcomes , 52– 101. doi:10.1093/acprof:oso/9780195368673.001.0001. Polat, F. ( 2003). Factors affecting psychosocial adjustment of deaf students. Journal of Deaf Studies and Deaf Education , 8, 325– 339. doi:10.1093/deafed/eng018. Google Scholar CrossRef Search ADS PubMed Shulman, N. ( 1975). Life-cycle variations in patterns of close relationships. Journal of Marriage and the Family , 37, 813– 821. doi:10.2307/350834. Google Scholar CrossRef Search ADS Smith, C. R. ( 1975). Residual hearing and speech production in deaf children. Journal of Speech, Language, and Hearing Research , 18, 795– 811. doi:10.1044/jshr.1804.795. Google Scholar CrossRef Search ADS Spencer, L. J., Tomblin, J. B., & Gantz, B. J. ( 2012). Growing up with a cochlear implant: Education, vocation, and affiliation. Journal of Deaf Studies and Deaf Education , 17, 483– 498. doi:10.1093/deafed/ens024. Google Scholar CrossRef Search ADS PubMed Sprecher, S., & Regan, P. C. ( 2002). Liking some things (in some people) more than others: Partner preferences in romantic relationships and friendships. Journal of Social and Personal Relationships , 19, 463– 481. doi:10.1177/0265407502019004048. Google Scholar CrossRef Search ADS Swenson, L. M., Nordstrom, A., & Hiester, M. ( 2008). The role of peer relationships in adjustment to college. Journal of College Student Development , 49, 551– 567. doi:10.1353/csd.0.0038. Google Scholar CrossRef Search ADS Tobey, E. A., Geers, A. E., Sundarrajan, M., & Shin, S. ( 2011). Factors influencing speech production in elementary and high-school aged cochlear implant users. Ear and Hearing , 32, 27S– 38S. doi:10.1097/AUD.0b013e3181fa41bb. Google Scholar CrossRef Search ADS van der Linden, D., Scholte, R. H., Cillessen, A. H., te Nijenhuis, J., & Segers, E. ( 2010). Classroom ratings of likeability and popularity are related to the Big Five and the general factor of personality. Journal of Research in Personality , 44, 669– 672. doi:10.1016/j.jrp.2010.08.007. Google Scholar CrossRef Search ADS © The Author(s) 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: firstname.lastname@example.org.
The Journal of Deaf Studies and Deaf Education – Oxford University Press
Published: Jan 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.
All for just $49/month
Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.
Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.
It’s easy to organize your research with our built-in tools.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera