Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Mar 22.
Published in final edited form as: Emotion. 2016 Nov 7;17(2):191–195. doi: 10.1037/emo0000234

Adult Age Differences in the Interpretation of Surprise Facial Expressions

Michael M Shuster 1, Joseph A Mikels 1, Linda A Camras 1
PMCID: PMC5864289  NIHMSID: NIHMS820121  PMID: 27819446

Abstract

Research on adult age differences in the interpretation of facial expressions has yet to examine evaluations of surprise faces, which signal that an unexpected and ambiguous event has occurred in the expresser’s environment. The present study examined whether older and younger adults differed in their interpretations of the affective valence of surprise faces. Specifically, we examined older and younger participants’ evaluations of happy, angry, and surprise facial expressions. Based on age-related changes in the processing of emotional information, we predicted that older adults would evaluate surprise faces more positively than younger adults. The results indicated that older adults interpreted surprise faces more positively than their younger counterparts. These findings reveal a novel age-related positivity effect in the interpretation of surprise faces, suggesting that older adults imbue ambiguous facial expressions – that lack either positive or negative facial actions – with positive meaning.

Facial expressions of surprise convey an individual's attempt to understand the presence of an unexpected yet highly salient event in their environment (Hortsmann, 2006). Younger adults have consistently been shown to interpret affectively ambiguous surprise faces as negative (Neta, Norris, & Whalen, 2009; Neta, & Whalen, 2010). However, as the processing of affective information shifts across the adult life span toward positivity (Carstensen & Mikels, 2005), there is good reason to suppose that the negativity bias in the processing of surprise faces does not extend into later adulthood. The present study thus examined whether older adults interpret surprise faces more positively than their younger counterparts.

Facial expressions of other individuals are salient cues that help us interpret social behavior and decide appropriate reactions to a situation. Although some expressions are clear in the emotional valence that they communicate (e.g., anger, fear, etc.), other expressions are ambiguous and can be interpreted as reflecting either a positive or a negative valence (Kim, et al., 2004; Leppänen, Milders, Bell, Terriere, & Hietanen, 2004; Neta & Whalen, 2011; Said, Sebe, & Todorov, 2009). Facial expressions can be ambiguous by either displaying a mix of positive and negative affect, or by displaying an absence of clearly positive and negative affect as in the case of surprised or neutral expressions.

Relative to neutral expressions, the presence of a surprise expression signals that a meaningful event has transpired in the environment (Hortsmann, 2006; Meyer, Reisenzein, Schützwohl, 1997). The surprise facial expression reflects an adaptive initial reaction to an unexpected event that has interrupted one’s actions and reoriented their attention to enhance responses to an event deviating from their expectations. From a functionalist perspective, the raised brows and widened eyes of a surprise expression serve to enhance the visual field to better update an individual’s schemas regarding important aspects of the environment (Shariff & Tracy, 2011). Surprise faces are evaluated as indicating that a person’s actions have been interrupted and that more information is needed prior to resuming action (Hortsmann, 2003). Most importantly, surprise faces are inherently neither positive nor negative – but positive or negative interpretations have been shown to depend on the presence of a disambiguating context (Kim, Somerville, Johnstone, Alexander, & Whalen, 2003; Kim et al., 2004). Surprise expressions have thus been useful in research examining interpretive biases.

A negativity bias in interpretations of surprise expressions has been found in children, adolescents, and younger adults (e.g., Neta et al., 2009; Tottenham et al., 2013). Although there was variability across individual ratings of surprise expressions, younger adult participants evaluated affectively ambiguous faces as more negative than positive (Neta et al., 2009; Neta & Whalen, 2010). Furthermore, oddball-type paradigms demonstrated that surprise expressions are processed more similarly to negative anger expressions compared to positive happy expressions in younger adults (Neta, Davis, & Whalen, 2011). Altogether, Neta and colleagues (2011) suggest that younger adults interpret surprised faces as being more negative than positive. However, due to motivational shifts toward emotion regulation goals in later life (see e.g., Charles & Carstensen, 2007), it is likely that developmental differences in the processing of surprise faces may be observed in late adulthood.

As proposed by socioemotional selectivity theory (SST; Carstensen, 2006), older adults are more motivated than their younger counterparts to optimize emotional wellbeing due to their constrained time horizons. This motivation leads older adults to process either positive information to a greater extent, or negative information to a lesser extent, than younger adults. The positivity effect–the age-related preference for positive as opposed to negative material in information processing (Carstensen & Mikels, 2005)–is considered to be motivated and volitional in nature; this effect is amplified when people are free to process information as they like, for instance, in unconstrained rather than constrained information processing tasks (Reed, Chan, & Mikels, 2014). However, the bulk of this research has used unambiguous positive and negative emotional material.

Researchers only recently have examined adult age differences in the interpretations of ambiguous information. For instance, compared to younger adults, older adults were found to generate less negative resolutions to emotionally ambiguous scenarios (Mikels & Shuster, 2016). Kellough & Knight (2012) presented older and younger adults with facial expressions representing a blend of discrete positive and negative emotions. Relative to younger adults, older adults provided more positive evaluations for the blended expressions. Blends of positive and negative expressions simultaneously display a partial smile in addition to components of sadness, anger, or fear. These blends display a mix of positive and negative affect and are thus fundamentally different from expressions of surprise. Surprise faces convey a discrete emotional reaction, which uniquely signals that an unexpected event has occurred and that the expresser is vigilantly trying to assess the situation (Hortsmann, 2003). Unlike blended expressions, surprise faces do not contain muscle actions that are exclusive to expressions of discrete positive or negative emotions. Studies examining the visual processing of facial expressions during emotion identification tasks revealed that older adults tend to divert attention from the eyes in favor of gazing at the mouth (e.g., Murphy & Isaacowitz, 2010). As such, older adults’ relatively more positive evaluations of blended expressions may result from perceptual biases rather than an interpretive bias.

The present study utilized surprise faces rather than blends in order to determine whether older relative to younger adults provide more positive evaluations of facial expressions that lack facial actions exclusively associated with positive or negative affect. To do so, we used a modified version of the facial evaluation task of Neta et al. (2009), where older and younger adults evaluated the valence of happy, angry, and surprise expressions. Based on previous findings regarding age differences in the processing of ambiguity, we predicted that, relative to younger adults, older adults would evaluate the surprise expressions as being more positive. For exploratory purposes, facial electromyography was also collected to examine potential relations between affective evaluations and facial responses.

Method

Participants

Thirty-one younger adults and 32 older adults were recruited (see Table 1 for participant characteristics). This sample size is comparable to previous research that examined age differences in the interpretation of ambiguous situations (Mikels et al., 2016) and studies examining relations between evaluative biases and fEMG activity (Neta et al., 2009). Older adults were compensated for their participation monetarily; younger adults were compensated with payment or course credit.

Table 1.

Participant Characteristics by Age Group.

Younger Older Statistic
M (SD) M (SD) t p
Age (in years) 20.93 (3.23) 73.56 (7.44)
Sex 73.3% F, 26.7% M 71.9% F, 28.1% M
Education (in years) 13.37 (3.25) 15.66 (3.26) −2.77 .008
Socioeconomic Level 2.84 (1.10) 2.87 (0.94) −0.14 .828
Vocabulary (WAIS-IV) 32.42 (8.92) 41.84 (9.62) −4.03 <.001
Coding (WAIS-IV) 83.68 (12.78) 60.75 (15.22) 6.12 <.001
Digit Span (WAIS-IV) 28.16 (4.97) 25.87 (5.76) 1.67 .099
Overall Valence Ratings 0.58 (0.69) 1.20 (0.69) −3.50 <.001
Angry Valence Ratings −2.04 (0.34) −2.21 (0.44) 1.72 .091
Happy Valence Ratings 2.14 (0.31) 2.30 (0.44) −1.61 .114
Surprise Valence Ratings −0.16 (0.53) 0.59 (0.71) −4.77 <.001
Corrugator Activity −0.05 (0.17) −0.24 (0.26) 3.43 .010
Zygomaticus Activity −0.02 (0.05) −0.06 (0.06) 2.91 .005
Open in a new tab

Note. Sex: F = Female, M = Male; Socioeconomic Level: on a scale of 1–5 (with 1 indicating “lower income” and 5 indicating “upper income”); Vocabulary from the Wechsler Adult Intelligence Scale (WAIS–IV; Wechsler, 2008): maximum score = 57; Digit-Symbol Coding from the WAIS-IV: maximum score = 135; Digit Span from the WAIS-III: maximum score = 48; Valence Ratings were measured on a scale of −3 (very negative) to +3 (very positive).

Stimuli

Following Neta and colleagues’ (2009) methodology, images of facial expressions of anger, happiness, and surprise were selected from 9 male and 9 female models from the NimStim database of emotional facial expressions that were validated as having high rates of intra-participant agreement in emotion identification tasks (Tottenham et al., 2009).

Facial Electromyography Measures

Facial electromyography (fEMG) measured participants’ facial responses to the images (Larsen, Norris, & Cacioppo, 2003). Pairs of 4 mm Ag/AgCl electrodes were attached to the corrugator (brow) and zygomaticus (cheek) muscle sites according to Fridlund and Cacioppo’s (1986) guidelines. Facial muscle activity was recorded at a sampling rate of 1 kHz with an integrated wireless system and software package (Biopac MP150, AcqKnowledge; Biopac Systems, Goleta, CA). Measures were collected across the entire task. The fEMG data were processed according to the protocol used in previous physiological examinations of affect (e.g., Mikels & Shuster, 2016).

Assessments of Cognitive Ability

Three standard WAIS-IV (Weschler, Coalson, & Raiford, 2008) measures of cognitive functioning were included to compare older and younger adults. Vocabulary required participants to provide brief verbal definitions of word lists that increased in difficulty. Coding measured participants’ processing speed by requiring them to match symbols that corresponded to digits as quickly as possible for two minutes. Digit Span measured short-term memory (STM), having participants remember and repeat strings of digits. See Table 1 for means and standard deviations.

Procedure

After consent, participants were fitted with facial EMG sensors placed over their corrugator and zygomaticus muscles. Participants completed a five-minute acclimation period to accustom them to the sensors. Next, participants were informed they would be viewing and evaluating a series of facial expressions that would be presented for brief durations. Each of the 54 images was presented twice over two separate runs in random order for a total of 108 trials. Images were presented on a computer screen with a white background, one at a time, for one second each. Before each expression, participants were presented with a black fixation cross on a white screen for six seconds followed by a white screen with a red fixation cross for 500ms (to help the participants orient to the upcoming image). After each image, participants rated the expression’s valence using labeled keyboard keys. The labeled keys represented a 6-point scale ranging from −3 (very negative) to +3 (very positive). Lastly, the fEMG sensors were removed from the participants who were then administrated the cognitive tasks and demographic survey.

Results

To examine age differences in the valence ratings of the three facial expressions, a 2 (age group) x 3 (expression) mixed measures ANOVA was conducted. A main effect of expression emerged indicating that valence ratings were significantly different across the three categories of facial expression, F(2, 122) = 1285.37, p< .001, ηp2= .955. Bonferroni corrected pairwise comparisons revealed that the average valence ratings for each expression were significantly different from the others. Angry expressions were rated the most negative (M= −2.13, SD= .40), happy expressions were rated the most positive (M= 2.22, SD= .39), and surprised expressions were rated in the middle (M= .21, SD= .73). Furthermore, a main effect of age emerged such that older adults (M= .23, SD= .32) rated facial expressions significantly more positively overall compared to younger adults (M= −.02, SD= .23), F(1, 61)= 12.25, p= 0.001, ηp2= .167.

Importantly, the interaction between age group and expression was significant, F(2, 122)= 14.91, p< .001, ηp2= .196. An independent samples t-test was conducted for each expression type to compare the valence ratings provided by older and younger adults. The valence ratings provided by younger (M= −2.04, SD= .34) and older (M= − 2.21, SD= .44) adults did not significantly differ for angry expressions, t(61)= 1.72, p= .09, d= .432. Similarly, for happy expressions the older (M= 2.30, SD= .44) and younger (M= 2.14, SD= .31) adults did not significantly differ on their valence ratings, t(55.25)= −1.61, p= .114, d= .420. However, older adults rated surprised facial expressions significantly more positively (M = 0.59, SD= 0.71) than did younger adults (M= −0.16, SD= 0.53), t(57.47)= −4.77, p< 0.001, d= 1.197. The above analyses were repeated with the inclusion of the three cognitive assessment scores as covariates to examine if age differences in valence ratings were possibly explained by age differences in cognitive performance. Including these scores did not change the pattern or significance of the results.

To examine differences in fEMG activity, a 2 (age group) x 3 (expression) mixed measures ANOVA was conducted for corrugator and zygomaticus separately. A main effect of age emerged for both corrugator and zygomaticus activity indicating that younger adults had higher levels of activity compared to older adults, F(1, 61)= 11.79, p= .001, ηp2= .162, and F(1, 61)= 40.57, p= .005, ηp2= .122 respectively (see Table 1). No other effects emerged.

Discussion

This study provides evidence for age differences in the evaluation of affectively ambiguous surprise faces. As predicted, relative to younger adults, older adults interpreted surprise expressions as being more positive. The valence ratings provided by older and younger adults did not significantly differ for angry or happy expressions. Thus, the age difference in affective interpretations of facial expressions was specific only to the surprise expressions. This finding provides an important boundary condition to the positivity effect in affective evaluations, such that only ambiguous surprise faces may be susceptible to age differences in interpretation. Such age differences could have inadvertent downstream consequences, insofar as interpretations of surprise faces in unpredicted situations can influence how people appraise their social partners’ intentions (Justyte & Schönenberg, 2014).

These findings are consistent with recent work on age differences in affective evaluations and interpretations of facial expressions that were mixed in terms of valence (Kellough & Knight, 2012). Our study’s use of surprise expressions rather than positive-negative blends extends Kellough and Knight’s (2012) findings into expressions of surprise that uniquely signal an adaptive emotional reaction that is initiated by the appraisal of uncertainty caused by changing aspects of the environment. Our findings also parallel research demonstrating that relative to the young, older adults form more positive trait impressions (e.g., health and trustworthiness) of neutral facial expressions (Zebrowitz, Franklin, Hillman, & Boc, 2013). Along with Zebrowitz et al., (2013), our findings extend the scope of the positivity effect into evaluations of socio-emotional stimuli that are neither positive nor negative in valence. Previous research on the positivity effect has been limited to using stimuli that were either positive, negative, or mixed valence (see: Reed, Chan, & Mikels, 2014). Due to our use of surprise faces (rather than mixed expressions), our findings suggest that older adults imbue facial expressions that are neither positive nor negative in terms of their valence with positive meaning.

Although the present investigation documented age differences in the interpretation of ambiguous facial expressions, there are limitations regarding generalization to real life contexts. Recent literature suggests that the context surrounding facial expressions may sometimes determine how those expressions are identified and that older (compared to younger) adults may be more reliant on context to correctly identify facial expressions (Noh & Isaacowitz, 2013). Thus, future research should consider the role of contextual factors surrounding ambiguous expressions to improve ecological validity. Regarding the exploratory fEMG data, our findings did not provide any additional insights, possibly due to lower facial reactivity among older adults. Future studies should include larger samples especially when examining physiological measures due to their greater variability.

Our investigation provides evidence for age differences in the interpretation of surprise faces that are ambiguous in terms of valence. Specifically, older adults rated the surprise faces as more positive in comparison to younger adults. Therefore, older versus younger adults may differentially appraise and interpret the faces of others during unpredicted situations, which could impact subsequent courses of action.

Acknowledgments

This research was supported by the National Institute on Aging, Grant R01-AG043533, & by the National Science Foundation, Grant SES-1139554 to JAM.

References

  1. Carstensen LL. The influence of a sense of time on human development. Science. 2006;312(5782):1913–1915. doi: 10.1126/science.1127488. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Carstensen LL, Mikels JA. At the intersection of emotion and cognition aging and the positivity effect. Current Directions in Psychological Science. 2005;14(3):117–121. [Google Scholar]
  3. Charles ST, Carstensen LL. Emotion regulation and aging. In: Gross JJ, editor. Handbook of Emotion Regulation. New York: Guilford Press; 2007. pp. 307–320. [Google Scholar]
  4. Fridlund AJ, Cacioppo JT. Guidelines for human electromyographic research. Psychophysiology. 1986;23(5):567–589. doi: 10.1111/j.1469-8986.1986.tb00676.x. [DOI] [PubMed] [Google Scholar]
  5. Horstmann G. The psychological refractory period of stopping. Journal of Experimental Psychology: Human Perception and Performance. 2003;29(5):965. doi: 10.1037/0096-1523.29.5.965. [DOI] [PubMed] [Google Scholar]
  6. Horstmann G. Latency and duration of the action interruption in surprise. Cognition & Emotion. 2006;20(2):242–273. [Google Scholar]
  7. Jusyte A, Schönenberg M. Threat processing in generalized social phobia: An investigation of interpretation biases in ambiguous facial affect. Psychiatry Research. 2014;217(1):100–106. doi: 10.1016/j.psychres.2013.12.031. [DOI] [PubMed] [Google Scholar]
  8. Kellough JL, Knight BG. Positivity effects in older adults’ perception of facial emotion: The role of future time perspective. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences. 2012;67(2):150–158. doi: 10.1093/geronb/gbr079. [DOI] [PubMed] [Google Scholar]
  9. Kim H, Somerville LH, Johnstone T, Alexander AL, Whalen PJ. Inverse amygdala and medial prefrontal cortex responses to surprised faces. Neuroreport. 2003;14(18):2317–2322. doi: 10.1097/00001756-200312190-00006. [DOI] [PubMed] [Google Scholar]
  10. Kim H, Somerville LH, Johnstone T, Polis S, Alexander A, Shin LM, Whalen PJ. Contextual modulation of amygdala responsivity to surprised faces. Journal of Cognitive Neuroscience. 2004;16(10):1730–1745. doi: 10.1162/0898929042947865. [DOI] [PubMed] [Google Scholar]
  11. Larsen JT, Norris CJ, Cacioppo JT. Effects of positive and negative affect on electromyographic activity over zygomaticus major and corrugator supercilii. Psychophysiology. 2003;40(5):776–785. doi: 10.1111/1469-8986.00078. [DOI] [PubMed] [Google Scholar]
  12. Leppänen JM, Milders M, Bell JS, Terriere E, Hietanen JK. Depression biases the recognition of emotionally neutral faces. Psychiatry Research. 2004;128(2):123–133. doi: 10.1016/j.psychres.2004.05.020. [DOI] [PubMed] [Google Scholar]
  13. Meyer WU, Reisenzein R, Schützwohl A. Toward a process analysis of emotions: The case of surprise. Motivation and Emotion. 1997;21(3):251–274. [Google Scholar]
  14. Mikels JA, Shuster MM. The interpretative lenses of older adults are not rose-colored—just less dark: Aging and the interpretation of ambiguous scenarios. Emotion. 2016;16(1):94–100. doi: 10.1037/emo0000104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Mikels JA, Shuster MM, Thai ST, Smith-Ray R, Waugh CE, Roth K, … Stine-Morrow EL. Messages that matter: Age differences in affective responses to framed health messages. Psychology and Aging. 2016;31(4):409–414. doi: 10.1037/pag0000040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Murphy NA, Isaacowitz DM. Age effects and gaze patterns in recognising emotional expressions: An in-depth look at gaze measures and covariates. Cognition and Emotion. 2010;24(3):436–452. [Google Scholar]
  17. Neta M, Whalen PJ. The primacy of negative interpretations when resolving the valence of ambiguous facial expressions. Psychological Science. 2010;21(7):901–907. doi: 10.1177/0956797610373934. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Neta M, Whalen PJ. Individual differences in neural activity during a facial expression vs. identity working memory task. Neuroimage. 2011;56(3):1685–1692. doi: 10.1016/j.neuroimage.2011.02.051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Neta M, Davis FC, Whalen PJ. Valence resolution of ambiguous facial expressions using an emotional oddball task. Emotion. 2011;11(6):1425. doi: 10.1037/a0022993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Neta M, Norris CJ, Whalen PJ. Corrugator muscle responses are associated with individual differences in positivity-negativity bias. Emotion. 2009;9(5):640. doi: 10.1037/a0016819. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Noh SR, Isaacowitz DM. Emotional faces in context: Age differences in recognition accuracy and scanning patterns. Emotion. 2013;13(2):238. doi: 10.1037/a0030234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Reed AE, Chan L, Mikels JA. Meta-analysis of the age-related positivity effect: Age differences in preferences for positive over negative information. Psychology and Aging. 2014;29(1):1. doi: 10.1037/a0035194. [DOI] [PubMed] [Google Scholar]
  23. Said CP, Sebe N, Todorov A. Structural resemblance to emotional expressions predicts evaluation of emotionally neutral faces. Emotion. 2009;9(2):260. doi: 10.1037/a0014681. [DOI] [PubMed] [Google Scholar]
  24. Shariff AF, Tracy JL. What are emotion expressions for? Current Directions in Psychological Science. 2011;20(6):395–399. [Google Scholar]
  25. Tottenham N, Phuong J, Flannery J, Gabard-Durnam L, Goff B. A negativity bias for ambiguous facial-expression valence during childhood: Converging evidence from behavior and facial corrugator muscle responses. Emotion. 2013;13(1):92. doi: 10.1037/a0029431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Tottenham N, Tanaka JW, Leon AC, McCarry T, Nurse M, Hare TA, … Nelson C. The NimStim set of facial expressions: Judgments from untrained research participants. Psychiatry Research. 2009;168(3):242–249. doi: 10.1016/j.psychres.2008.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Wechsler D, Coalson DL, Raiford SE. WAIS-IV: Wechsler adult intelligence scale. San Antonio, TX: Pearson; 2008. [Google Scholar]
  28. Zebrowitz LA, Franklin RG, Jr, Hillman S, Boc H. Older and younger adults' first impressions from faces: Similar in agreement but different in positivity. Psychology and Aging. 2013;28(1):202. doi: 10.1037/a0030927. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES