Introduction

Trust is essential for human cooperation to emerge both at a small scale, such as among family members and friends, and at a larger scale, such as among groups within a society. Thus, recognizing trustworthy partners who serve not only themselves but are likely to engage in reciprocity is of individual and communal benefit. Although it is well known that humans aim to maximize this benefit by gauging the looks and behaviors of others [1,2], the underlying mechanisms are still poorly understood. Moreover, it remains unclear whether and how vocal signals are relevant. Here, we sought to better understand their role and to explore what aspects of a speaker’s voice bias listener trust. Specifically, we investigated how expressed affect and speaker person characteristics (i.e., age, sex) shape trustworthiness perceptions and examined the acoustic patterns that explain their influence.

The vocal apparatus, that is the instrument from which the human voice emanates, is a complex system comprising a wind generator—the lungs, a string-like vibrator—the vocal folds, and a sequence of resonance chambers such as the pharynx, mouth and nasal cavities [3]. Relatively fixed structural aspects (e.g., vocal tract length, vocal fold size) as well as dynamic bodily activity (e.g., breathing, muscle tension) modulate the workings of the vocal apparatus and thus how the voice sounds. Resulting inter- and intra-individual differences can be quantified by a range of acoustic measures, also referred to as acoustic parameters, that describe a sound’s frequency, intensity, and duration characteristics.

A substantial body of research has demonstrated statistical relationships between certain acoustic profiles and the affective state that listeners attribute to a speaker [4–6]. For example, fundamental frequency (F0), which is perceived as pitch or voice melody, is higher for elation as compared with other emotions such as anger and sadness [4]. Additionally, there is evidence that voice acoustics predict listener impressions of more stable person characteristics [7–14] including the extent to which a speaker appears trustworthy. Specifically, a lower F0 mean in speech has been repeatedly linked to increased perceived kindness or trustworthiness [8,15–18]. Additionally, it has been shown that utterances rated as high and low on trustworthiness differ in F0 contour [19]. For example, in one study highly trustworthy pronunciations of the word “hello” were characterized by a rising F0 from the first to the second syllable. Moreover, when applied to other words, this rising F0 contour was found to influence listeners in choosing the more trustworthy of two voices [20].

However, not all research corroborates these observations and attests to a simple relationship between trustworthiness and F0. For example, in some hands F0 mean correlated positively with rated trustworthiness [20] and in other hands it was irrelevant [21]. Moreover, the direction of the F0 effect was shown to depend on context or the framing of the judgement task [22–25]. For example, in economic or mating-related contexts, voices with a higher F0 seemed more trustworthy than voices with a lower F0 irrespective of speaker sex. However, when probed more generally, impressions of trustworthiness appeared to differ by sex with low F0 female and high F0 male voices being preferred [23].

Although past investigations of voice acoustics offered insights into the mechanisms underpinning vocal trustworthiness, more research is needed on how acoustically derived information about the speaker is relevant. Specifically, it is of interest to determine the extent to which expressive and person characteristics, and thus potentially higher-order conceptual processes, shape trustworthiness impressions. Evidence that they might comes from the face processing literature. Research on dynamic face characteristics indicates that positive or happy expressions are rated as more trustworthy than neutral expressions, which are perceived as more trustworthy than negative expressions such as anger [26,27]. The role of stable face characteristics has been explored by manipulating person characteristics such as sex and age. Although women consistently seem more trustworthy than men [28,29], results for age are mixed. In the context of a trust game, participants expressed greater trust towards older as compared with younger individuals [30], yet their trustworthiness ratings did not map onto actual behavior in the game. Moreover, other work assessing trustworthiness impressions found them to be independent of age [31] or to favour younger over older adults [32].

Effects of both speaker expressive and person characteristics on perceived trustworthiness have been explained by the “emotion over-generalization” hypothesis, first conceived as metaphorical generalization by Secord [33,34]. According to this framework, emotional expressions are important for signalling probable behaviors and because positive affect is believed to increase the likelihood of a benevolent intent it purportedly increases trust. Additionally, the similarity between emotional expressions and person characteristics is thought to explain why the latter are co-opted when judging whom to trust [35–37]. For example, average female faces have thinner inner eyebrows and rounder jaws than average male faces and this, among other features, boosts perceptions of positive affect or surprise and therefore, so the hypothesis suggests, of trustworthiness. Support for this, albeit not entirely consistent [38], comes from a number of studies on faces [37,39].

To date, only a couple of studies considered expressive and person characteristics in the context of vocal first impressions [10,40,41] and none, that we could identify, examined trustworthiness. Hence, we sought to address this issue. We invited 20 speakers, differing in age and sex, to produce two sentences in accordance with specific expressive instructions (see below). Each expression was subjected to acoustic analysis and to a trustworthiness rating by a group of 80 listeners. Statistical analyses tested the following three predictions. (1) Expression valence was expected to correlate with perceived trustworthiness. As shown for faces, positive voices should be rated as more trustworthy than negative voices. We also pursued a possible, and not previously delineated, relation between the arousal and trustworthiness of vocal expressions. Besides valence, arousal, or a person’s level of energy, forms a major dimension in a basic affective space [42,43] and serves as a relevant cue to behavioral intentions. It may, therefore, also play a role in biasing trust. (2) A second hypothesis was that basic person characteristics like sex and age would modulate perceived trustworthiness. Again, this prediction derived from previous work on faces showing that person characteristics bias trust preferences [28,30]. In line with extant findings, we expected female voices to receive higher trustworthiness ratings than male voices. However, due to inconsistent results no directional predictions could be made for age effects. (3) Last, and in line with the emotion over-generalization hypothesis, we expected that the voice acoustics predicting higher perceived trustworthiness should also be the voice acoustics characterizing the expressive and person characteristics that are most trusted. In other words, if a low F0 mean is associated with high perceived trustworthiness, then the affective expression, sex, and age receiving the highest trustworthiness ratings should have a lower F0 mean than the affective expression, sex, and age receiving the lowest trustworthiness ratings.

Materials and methods

This research was approved by the Institutional Review Board of the National University of Singapore and conducted following established guidelines. Participants were informed about the study procedure and signed a consent form before commencing their participation.

Stimuli

The stimuli used in this study were recorded from 20 Singaporean native English speakers with acting experience and are available on the Open Science Framework (https://osf.io/j3hfg/?view_only=77db09592fe44a79837f61ba8146fdb8). Speakers included five young women with a mean age of 22.2 years (range 21–24), five young men with a mean age of 23.8 years (range 23–25), five older women with a mean age of 69.2 years (range 62–85), and five older men with a mean age of 63 years (range 45–79). The older individuals were recruited from a local amateur-acting association, whereas the younger individuals were recruited through campus advertising. Acting experience ranged from 1 to 30 years (mean 5.75, SD 3.8) for the older speakers and from 1 to 6 years (mean 2.5, SD 1.65) for the younger speakers.

Speakers attended individual recording sessions. They were given two English sentences (“The politician faced his audience.”, “The message reached its target.”) that the authors of this study deemed to be affectively neutral but that, depending on the context, could be potentially relevant for a range of affective experiences. They were asked to produce these sentences with a range of expressive intentions. These expressive intentions were introduced to generate utterances with varying intonations and were deemed potentially relevant for modulating perceived trustworthiness by the authors of this manuscript. The specific voice conditions included six emotional expressions (i.e., content, happy, proud, afraid, angry, and sad) and four conversational expressions (i.e., confident, stating, doubtful, and questioning). Additionally, we asked speakers to attempt to sound trustworthy, untrustworthy, and neutral. For each expressive intention, speakers were given an expression label along with a short sentence describing a scenario for them to interpret vocally (e.g., trustworthy—You are talking to a stranger whom you wish to trust in you.). Speakers were asked to warm-up their voice by reading the sentences aloud neutrally. Subsequently, they could tackle the individual expressive intentions in their preferred order and repeat attempts until they were satisfied with their portrayal. Once they were satisfied with their portrayal, recordings were made with a few repetitions from which we later picked the clearest exemplar for each speaker, expressive intention, and sentence (e.g., no microphone artefacts, clear pronunciation, identifiable expressive intention). Recordings were conducted in a sound attenuated chamber and digitized at a 16 bit/44.1 kHz sampling rate. They were processed off-line using Adobe Audition 2.0. Specifically, individual sentences were cut into separate files and their intensity was normalized at the root-mean-square value.

In a preliminary rating study, 30 individuals (15 women, average age of 22 years (SD 3.77)) listened to each expression (2 sentences x 13 expressions x 20 speakers = 520) and rated its perceived valence and arousal on 7-point scales ranging from very negative to very positive and from very calm to very aroused. Valence and arousal ratings were subjected to separate ANOVAs with Expression (13 levels) as a repeated measures factor. This produced a significant main effect of Expression for both the Valence (F[12,348] = 44.25, p < .0001, gη² = 0.481) and Arousal (F[12,348] = 74.99, p < .0001, gη² = 0.466) measures, indicating that there was sufficient variation in how positive/negative and calm/excited the stimuli sounded. A Pearson correlation analysis conducted on the mean rating values for each sentence and expression revealed a positive relation between valence and arousal (r(24) = 0.53, p = .005). Rating results are illustrated in Fig 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1.

Box and whisker plots illustrating perceived valence (top) and perceived arousal (bottom) for each expression for a by-item analysis that used data points for each speaker and expression averaged across listeners and sentences. Black dashed lines indicate the rating scale midpoint.

https://doi.org/10.1371/journal.pone.0210555.g001

Using Praat [45], we analysed the acoustic stimulus properties. Specifically, we measured eight parameters including sound duration (i.e., duration of the entire utterance), which may be considered a measure of speech rate as the number of syllables was kept constant across speakers and expressions (10 and 7 for sentences 1 and 2, respectively). The other parameters comprised harmonic-to-noise ratio (HNR; the periodicity of the signal; Harmonicity settings = 0.01/75/0.1/4.5, Time Step = 0.01 s, Minimum pitch = 75 Hz, Silence threshold = 0.1, Number of periods per window = 4.5), F0 mean (the lowest frequency band of an utterance; Time step: standard value 0.0, Pitch floor: standard value 75 Hz, Pitch ceiling: standard value 600 Hz), F0 range (the F0 difference between the lowest and the highest value in an utterance) and F0 standard deviation, intensity range (Intensity settings = 75/0/yes, Minimum Pitch = 75 Hz, Time Step = 0), jitter (i.e., short-term variability/perturbations in F0; Voice report settings = 0/0/75/600/1.3/1.6/0.03/0.45), and shimmer (i.e., short-term variability/perturbations in voice intensity; Voice report settings = 0/0/75/600/1.3/1.6/0.03/0.45).

Measurement values of all parameters were normalized (mean 0, SD 1) and subjected to a series of cumulative link mixed models (i.e., ordered logit regression models) using the clmm function from the Ordinal package (Christensen, 2018) in R (R Core Team, 2015) with its default settings. In a first set of models, valence was the dependent variable, the individual parameters were modelled as fixed main effect and the intercepts of Rater and Speaker were modelled as random effects. For valence, we observed, in order of effect size, a negative relationship with Duration (β = -0.514, SE = 0.018, Z = -28.38, p < .0001), HNR (β = -0.491, SE = 0.02, Z = -24.11, p < .0001), F0 SD (β = -.176, SE = 0.018, Z = -9.96, p < .0001), F0 mean (β = -.109, SE = 0.025, Z = -4.36, p < .001), Jitter (β = -0.079, SE = 0.018, Z = -4.37, p < .0001), and F0 range (β = -0.073, SE = 0.017, Z = -4.52, p < .0001). Given the small effect sizes, the Jitter and F0 range effects are likely negligible. There was a small positive relationship with Intensity Range (β = 0.163, SE = 0.021, Z = 7.88, p < .0001) and the effect with Shimmer was non-significant (β = -0.006, SE = 0.018, Z = -0.31, p = .758).

Analyzing arousal in an analogous way revealed a positive relationship with Intensity range (β = 0.724, SE = 0.021, Z = 34.17, p < .0001) and F0 mean (β = 0.455, SE = 0.026, Z = 17.25, p < .0001) and a negative relationship with HNR (β = -0.626, SE = 0.02, Z = -30.68, p < .0001), Duration (β = -0.339, SE = 0.018, Z = -18.89, p < .0001), F0 SD (β = -0.333, SE = 0.018, Z = -18.47, p < .0001), F0 range (β = -0.227, SE = 0.016, Z = -14.04, p < .0001), and with Shimmer (β = -0.225, SE = 0.018, Z = -12.49, p < .0001). The Jitter effect was non-significant (β = -0.011, SE = 0.018, Z = -0.62, p = .536).

Last we explored whether and how the two sex and age groups differed on the six acoustic parameters explored here. To this end, we fitted six linear mixed effect models using the lme4 package [46] in R [47]. An acoustic parameter served as the dependent variable, Speaker Age and Speaker Sex served as fixed main effects (in consideration of group size, the interaction between Speaker Age and Speaker Sex was not modelled), and the intercepts of Speaker and Expression served as random effects. Compared to older voices, younger voices had, in order of effect size, a shorter utterance Duration (β = -0.835, SE = 0.172, t = -4.85, p < .001), a lower HNR (β = -0.829, SE = 0.254, t = -3.27, p = .004), a smaller F0 SD (β = -0.666, SE = 0.214, t = -3.11, p = .005), a larger Intensity range (β = 0.633, SE = 0.274, t = 2.27, p = .035), a lower F0 mean (β = -0.533, SE = 0.159, t = -3.36, p = .003), and a marginally smaller F0 range (β = -0.354, SE = 0.184, t = -1.92, p = .069). There was no age effect on Jitter (β = -0.061, SE = 0.235, t = -0.26, p = .798) and Shimmer (β = -0.351, SE = 0.228, t = -1.54, p = .14). Compared to female voices, male voices had a lower F0 mean (β = -1.378, SE = 0.159, t = -8.68, p < .0001), more Jitter (β = 0.693, SE = 0.235, t = 2.95, p = .008), more Shimmer (β = 0.545, SE = 0.228, t = 2.38, p = .027), and a marginally smaller F0 range (β = -0.332, SE = 0.184, t = -1.8, p = .087). All other effects were non-significant (Duration, β = -0.062, SE = 0.172, t = -0.36, p = .72; HNR, β = -0.069, SE = 0.254, t = -0.27, p = .788; F0 SD, β = -0.322, SE = 0.214, t = -1.5, p = .147; Intensity range, β = 0.415, SE = 0.274, t = 1.51, p = .146).

Results

Trustworthiness and speaker affective expression

First, we examined the relationship between perceived trustworthiness—our ordinal dependent variable—and the affective rating of each item averaged across all raters (see Methods). Specifically, we fitted a cumulative link mixed model (i.e., ordered logit regression model) using the clmm function in the Ordinal package [49] in R [47]. Our model included Trustworthiness as the dependent variable, either normalized Valence or Arousal (normed item values, mean 0, SD 1) as the fixed effect and the intercepts of Listener (40 levels) and Speaker (20 levels) as random effects. This analysis was conducted for young listeners only because only young listeners participated in the preliminary rating study from which we derived affective item norms.

As illustrated in Fig 2 (top row), perceived trustworthiness was positively related to Valence (β = 0.56, SE = 0.014, Z = 40.2, p < .0001) but unrelated to Arousal (β = 0.017, SE = 0.014, Z = 1.211, p = .226). Control analyses indicated that this pattern replicated across the two sentences for Valence (sentence 1: β = 0.583, SE = 0.02, Z = 29.15, p < .0001; sentence 2: β = 0.554, SE = 0.02, Z = 27.04, p < .0001) but was inconsistent for arousal (sentence 1: β = 0.131, SE = 0.02, Z = 6.48, p < .0001; sentence 2: β = -0.11, SE = 0.02, Z = -5.49, p < .0001).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Trustworthiness rating results.

Top row—The scatter plots illustrate the relation between valence and trustworthiness (left) and arousal and trustworthiness (right). Colours represent individual speakers (N = 20) in a consistent manner across the two graphs. Data points reflect trustworthiness averages for each expression and sentence computed across listeners and are summarized with a regression line. Bottom row–Plots illustrating the interaction of speaker and listener characteristics for sex (left) and age (right), respectively. Female and younger speakers are represented by red dots, whereas male and older speakers are represented by grey dots. Data points reflect trustworthiness averages for speaker computed across expressions, sentences and listeners. Median values are marked by a black cross.

https://doi.org/10.1371/journal.pone.0210555.g002

Acoustic features of trustworthiness

As for the stimulus valence/arousal analyses reported in the methods section, we entered the normalized acoustic parameters into separate CLM model as the fixed main effect. Trustworthiness ratings served as the dependent variable and the intercepts of Listener and Speaker were random effects. The results are illustrated in Fig 3 along with the stimulus properties detailed in the Methods section.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Illustration of the relationship between acoustic measures and other stimulus properties.

Normalized vocal parameters (y-axes) are plotted as a function of trustworthiness ratings, valence, arousal, and speaker characteristics (x-axes). The individual points in the scatter plots represent the 13 expressions for the 2 sentences produced by each speaker. These points are summarized by a regression line for each speaker. The error bars in the bar-graphs represent the 95% confidence intervals.

https://doi.org/10.1371/journal.pone.0210555.g003

In order of effect size, higher trustworthiness was associated with a shorter utterance Duration (β = -.61, SE = .011, Z = -54.71, p < .0001), a lower HNR (β = -.429, SE = .012, Z = -35, p < .0001), a lower F0 mean (β = -0.402, SE = 0.015, Z = -26.28, p < .0001), a smaller F0 SD (β = -0.136, SE = 0.011, Z = -12.62, p < .0001), more Jitter (β = 0.104, SE = 0.011, Z = 9.28, p < .0001), more Shimmer (β = 0.102, SE = 0.011, Z = 9.14, p < .0001), a smaller F0 range (β = -0.053, SE = 0.009, Z = -5.43, p < .0001), and a larger Intensity range (β = 0.025, SE = 0.012, Z = 1.99, p = .046). Given effect sizes, F0 range and Intensity range effects are likely negligible.

We again checked these results by conducting separate analyses for each sentence. The results were comparable for sentence 1 (Duration, β = -.710, SE = .017, Z = -40.9, p < .0001; HNR, β = -.492, SE = .019, Z = -25.65, p < .0001; F0 mean, β = -.447, SE = .022, Z = -20.08, p < .0001; F0 SD, β = -.197, SE = .015, Z = -13.11, p < .0001; Jitter, β = .278, SE = .019, Z = 14.74, p < .0001; Shimmer, β = .149, SE = .019, Z = 7.87, p < .0001; F0 range, β = -.055, SE = .014, Z = -3.91, p < .0001; Intensity range, β = .037, SE = .018, Z = 2.06, p = .039) and for sentence 2 with the exception that the Jitter effect was non-significant (Duration, β = -.549, SE = .016, Z = -34.77, p < .0001; HNR, β = -.425, SE = .018, Z = -23.74, p < .0001; F0 mean, β = -.445, SE = .022, Z = -20.1, p < .0001; F0 SD, β = -.157, SE = .016, Z = -9.57, p < .0001; Jitter, β = .018, SE = .015, Z = 1.16, p < .0001; Shimmer, β = .047, SE = .014, Z = 3.2, p = .001; F0 range, β = -.069, SE = .014, Z = -4.76, p < .0001; Intensity range, β = .052, SE = .018, Z = 2.89, p < .0001).

Discussion

This study examined whether and in what way a speaker’s expressive and person characteristics shape listener trust.

Looking at expressive characteristics, we found a positive relationship between the valence of vocal expressions and rated trustworthiness. More positive voices were associated with increased perceived trustworthiness across our two stimulus sentences and across all speakers for the range of expressions that each speaker produced (Fig 2). Thus, we successfully replicate a phenomenon previously reported for faces [26,27]. Moreover, by providing an effect size estimate of 0.56 units on a 7-point rating scale, we highlight a potential relevance for this phenomenon in everyday life.

Notably, the arousal of vocal expressions predicted perceived trustworthiness inconsistently with a positive relation for one (“The politician faced his audience.”) and a negative relation for the other (“The message reached its target.”) sentence suggesting substantial dependence on verbal context and potential subtle differences in affective biases between the sentences. Moreover, a larger intensity range was, of all acoustic parameters, most strongly associated with arousal but had the smallest predictive value for the trustworthiness scores. Thus, it seems that the role of arousal in vocal trust may be negligible. Moreover, although valence and arousal conceptually overlap as indicated by a large correlation of rating results, they clearly dissociate in their significance for emerging first impressions.

The present study examined the role of person characteristics for trust from voices by looking at speaker sex and age effects. Extending evidence from faces [28,29], vocal recordings of women were rated as more trustworthy than those of men. Although this pattern showed consistently across the two sentences used here (betas = -.43, -.28), it was significant for only one of them, that is when speakers said “The politician faced his audience.” but not when they said “The message reached its target.”. Thus, although the female trustworthiness advantage may be a fairly robust phenomenon, its presence is moderated by other variables. For example, its strength may change as a function of whether trust concerns truthfulness or capability. In line with this, prevailing stereotypes characterize women as warmer but less competent than men [50].

Unlike our speaker sex effect, our speaker age effect showed consistently across the two sentences. Additionally, it was larger (β = 0.6) suggesting a greater relevance of age than sex in everyday trust. The age effect direction was both in conflict and in agreement with past research, which has provided a mixed pattern so far. Specifically, it was in conflict with work by Bailey and colleagues who found that older adult faces elicited higher trustworthiness ratings than younger adult faces [30]. However, those ratings were obtained after participants engaged in a trust-game with older and younger players and may, hence, have been context specific. Moreover, the age of the stimulus faces was left unspecified and may have been closer to middle age than to old age. This possibility agrees with research on political voting showing a preference for candidates in their 40s and 50s over younger and older candidates [15].

The present results corroborate evidence that younger faces appear more trustworthy than older faces [32] and that perceived vocal babyishness positively predicts ratings of warmth, honesty, and kindness [40]. Additionally, they converge with a more general age bias and a “youth ideal” that appeals on multiple levels and that readily engages processing resources. For example, research has shown that compared to expressions from older adults, expressions from younger adults more readily capture attention [51] and evoke better nonverbal perception [6]. Additionally, in surveys, younger individuals are seen as having better interpersonal skills as well as greater potential for professional development [52].

By examining both younger and older, female and male listeners, the present study also explored a potential role of peer status for perceived trustworthiness. Previous research identified own-age and own-sex preferences. Interpersonal attitudes have been characterized as more positive towards own as compared to other age individuals [52,53]. Additionally, children develop an own-sex bias that weakens after adolescence, but nevertheless persists into adulthood [54]. Together, this evidence raises the possibility that own-age and own-sex individuals are perceived as more trustworthy than those of another age and sex.

In line with prior work on facial age [31], however, the present study rejects this possibility in the context of vocal age. Younger and older listeners showed a similar age bias suggesting that youth is more relevant than peer status when it comes to trust. Yet, there was some indication for a role of peer effects in modulating sex-based preferences. Specifically, the preference for female over male voices tended to be larger in women than in men. Although awaiting replication, this tendency agrees with a recent study presenting untrustworthy, neutral, and trustworthy female and male faces [55]. Compared to men, women rated trustworthy faces as more trustworthy especially when they were female.

Both speaker expressive and person characteristics shaped voice acoustics and, through this, biased perceptions of trustworthiness. To see how, we measured voice acoustics and linked them to both the trustworthiness ratings and to the speaker variables (i.e., expression, age, sex). In doing so, we followed previous work on emotion expression [4] and examined a range of parameters describing frequency, intensity, and temporal sound properties.

Replicating earlier results [15–18], F0 mean negatively predicted perceived trustworthiness. Although this relationship has not been shown consistently [20,21], the present demonstration across a substantial speaker range gives additional credence. Apart from F0 mean, other parameters emerged as relevant. Specifically, ratings of trustworthiness were significantly higher for vocalizations with a shorter duration, a lower HNR, a smaller F0 SD, more shimmer, and–but for one sentence only–more jitter. Interestingly, duration (β = -.61) and HNR (β = -.43) turned out to be the strongest predictors with an effect size larger than that of other parameters including F0 mean (β = -.4), which has been the focus of research so far.

Looking at the acoustics of speaker affective expression, we found both overlap and dissociation with those of trustworthiness. Like trustworthy vocalizations, more positive vocalizations were associated with a faster speech rate, a lower HNR, a smaller F0 SD, and a lower F0 mean. Unlike trustworthy vocalizations, however, they showed a larger intensity range, a very small and negative relationship with jitter, and a non-significant relationship with shimmer. As for speaker affective expression, person characteristics produced mixed results. Compared to older voices, younger voices had a faster speech rate, a lower HNR, a smaller F0 SD, and a lower F0 mean, but also–contrary to prediction–a larger intensity range and a marginally smaller F0 range. Compared to male voices, female voices had a higher F0 mean, less jitter, and less shimmer–all results in opposition with the acoustics of trustworthiness.

Overall, the present study offers some support for the emotion over-generalization hypothesis [39,37], but see [38]. Perceived speaker valence, the most relevant dimension of positive affect positively predicted perceived speaker trustworthiness. Moreover, speech rate, HNR, F0 mean, and F0 SD which most strongly predicted valence, also most strongly predicted trustworthiness and, importantly, their prediction was in the same direction. Additionally, speaker age effects largely aligned with the emotion overgeneralization hypothesis. Young voices were more trusted than old voices and speech rate, HNR, F0 mean and F0 SD differences overlapped with the acoustic differences between positive and negative voices. Yet, for other parameters the patterns diverged between positive affect on the one hand and trustworthiness and age on the other hand. Additionally, speaker sex effects offered no support for the emotion overgeneralization hypothesis suggesting that positive affect is limited in explaining first impressions of trustworthiness. Instead, these impressions seem to emerge multi-causally. Some of their mechanisms may be largely perceptual and ‘hard-wired’ and tap on the biological convergence of expressive and person characteristics. However, others may depend on the more flexible, higher-order conceptual knowledge that individuals acquire across their life-time (e.g., stereotypes).

Although the present study provides insights into how vocal expressive and person characteristics shape trustworthiness impressions, it also raises questions for future research. One such question concerns the speaker sex and age effects observed here. The current speaker sample, albeit comparable to much published work [15–25], was small and one must worry that outliers biased overall effects and between speaker comparisons. We addressed this issue by controlling for individual speakers in our statistical analysis and by exploring whether effects emerge consistently for speakers across, as well as within, sex and age groups (Fig 2). As this was the case, we are confident in our results. Nevertheless, one may wish to replicate them with a larger and more representative sample.

As second question is how age affects both the vocal conveyance of trustworthiness and its perception. Looking at vocal production, it is still unclear how ageing changes the voice and the mechanisms of vocal expression. For example, although there is evidence that F0 mean increases in older adults [56] this evidence is inconsistent [57] and may interact with generational or cultural differences in life-style (e.g., smoking, singing) [57,58]. Additionally, it would be interesting to consider age differences in the motivation and ability to pose expressions and to examine spontaneous utterances as a function of age. Looking at vocal perception, the sensory, cognitive and emotional changes associated with ageing should be explored for their relevance for trustworthiness perception. For example future work may consider how age related hearing loss, decreasing executive functions [6], or a greater emphasis on positive experiences [59] may boost perceptions of trustworthiness in older adults.

Last, it would be important to pursue the role of voices for perceiving trustworthiness in everyday life. One issue is whether perceptions are misguiding or represent useful biases in situations with limited information. Clearly, we cannot be exceedingly accurate in judging trustworthiness from nonverbal cues only [1], but see [60]. Yet, there is empirical data suggesting that such judgements may not be entirely off the mark. Trustworthiness appears to have a stable personality basis [61] and observers appear to gauge some rudimentary aspect of this basis from faces more accurately than would be expected by chance [62–64]. Thus, one next step would be to explore the validity of trust impressions from the voice. Additionally, it would be relevant to examine how impressions inform actual behaviors.

Conclusions

To conclude, the present study pursued voice-based trustworthiness impressions from a fresh angle. In addition to exploring the sound properties of trustworthiness, it examined the role of vocal expressive and person characteristics and linked those back to relevant voice acoustics. In doing so, it produced original evidence that positive voices sound more trustworthy than negative voices. Additionally, it delineated speaker age and sex effects according to which trustworthiness is higher for younger as compared with older and female as compared with male individuals. Last, this study highlighted speech rate, HNR, F0 mean, and F0 SD as trust’s most important acoustic dimensions and showed that they map onto a subset of expressive and person characteristics providing some, albeit imperfect, support for the emotion overgeneralization hypothesis. Notably, a single acoustic parameter—fast speech—was as if not more important than speaker affect, age, and sex in predicting listener trust thus underlining that trust likely arises from both simple perceptual as well as higher-order conceptual processes.