Individual differences in the physical embodiment of care: Prosocially oriented women respond to cuteness by becoming more physically careful PDF

Preview

Summary

Emotion © 2012 American Psychological Association 2013, Vol. 13, No. 1, 151–158 1528-3542/13/$12.00 DOI: 10.1037/a0029259 Individual Differences in the Physical Embodiment of Care: Prosocially Oriented Women Respond to Cuteness by Becoming More Physically Careful Gary D. Sherman Jonathan Haidt Harva...

Description

Emotion 2013, Vol. 13, No. 1, 151–158

© 2012 American Psychological Association 1528-3542/13/$12.00 DOI: 10.1037/a0029259

Individual Differences in the Physical Embodiment of Care: Prosocially Oriented Women Respond to Cuteness by Becoming More Physically Careful Gary D. Sherman

Jonathan Haidt

Harvard University

University of Virginia

Ravi Iyer

James A. Coan

University of Southern California

University of Virginia

Prosocially oriented individuals tend to respond to care-relevant stimuli in a highly embodied manner. Research on facets of prosocial orientation—such as empathy—and embodiment has focused on processes triggered by the perception of others’ distress or pain. We suspect that the predisposition among prosocially oriented individuals toward having embodied responses to care-relevant stimuli might be more extensive. We tested the specific hypothesis that prosocial orientation would predict the likelihood of responding to cuteness (an understudied care stimulus that does not involve overt distress) with the physical embodiment of care: increased physical carefulness. In 2 studies, for prosocially oriented women only, cuteness elicited greater physical carefulness in a manual precision task. For such women, the elevated state of care elicited by cuteness cues is not only a coordinated set of feelings and motives but it is also a physically embodied state characterized by heightened carefulness in one’s physical movements. Keywords: care, embodiment, cuteness, prosocial orientation, empathy

Emotions are not just mental events; the effects of affective stimuli on perceivers are often distributed across multiple bodily systems. Traditionally, the study of the bodily component of emotion has focused on changes in either autonomic physiology (e.g., heart rate deceleration when viewing disgust images; Stark, Walter, Schienle, & Vaitl, 2005) or facial expression (e.g., zygomaticus major activity when viewing pleasant film clips; Ekman, Friesen, & Ancoli, 1980). But affective stimuli can also initiate more subtle changes in the motor system. For example, positively valenced stimuli can prime or facilitate approach behavior (e.g., pulling a lever; Chen & Bargh, 1999), suggesting that emotional stimuli can trigger preparations for emotion-congruent action. Someone who responds to an affective stimulus with this sort of covert, emotion-congruent change in the motor system can be said to be having an embodied response (Niedenthal, 2007). Although studies on the influence of affective stimuli on the motor system have primarily focused on general dimensions of emotion (e.g., valence and arousal) and on broad changes in the

motor system (e.g., approach vs. avoid), there may be more nuanced, emotion-specific embodiment effects. Indeed, Frijda (1986) conceptualized an emotion as a “felt mode of action readiness” (p. 238). We have targeted an understudied affective stimulus that may be embodied in this way: cuteness. The perception of someone or something as cute— usually because it possesses infantile characteristics—is known to trigger the phenomenological and motivational correlates of care (e.g., feelings of tenderness, desires to protect; Alley, 1983; Lorenz, 1950/1971; McCabe, 1988). We suspect that cuteness may also elicit the physical embodiment of care—actual carefulness in one’s behavior. We previously found support for this hypothesis: viewing cute images (baby animals) improved performance on the children’s game “Operation,” a task that requires physical carefulness to manipulate tweezers to remove small objects from small compartments (Sherman, Haidt, & Coan, 2009). Critically, this effect was observed despite the fact that the two conditions were matched in affective positivity and the manipulation did not produce consistent changes in physiological arousal (e.g., heart rate, skin conductance). Thus, it is unlikely that this effect was due to a general dimension of emotion such as valence or arousal. In our investigation, we expanded on this initial demonstration to understand how this embodied cuteness– carefulness effect is related to individual differences. Specifically, we tested the hypothesis that possessing a prosocial orientation—tending to prioritize the needs and welfare of others (Côté et al., 2011)—would predispose one to having an embodied response to cuteness and, therefore, to showing the cuteness– carefulness effect.

This article was published Online First August 13, 2012. Gary D. Sherman, Kennedy School of Government, Harvard University; Jonathan Haidt and James A. Coan, Department of Psychology, University of Virginia; Ravi Iyer, Department of Psychology, University of Southern California. We thank Kevin Eady for programming the cursor tracing task. Correspondence concerning this article should be addressed to Gary D. Sherman, Harvard Kennedy School of Government, 79 JFK Street, Mailbox 126, Cambridge, MA 02138. E-mail: [email protected] 151

SHERMAN, HAIDT, IYER, AND COAN

152

This hypothesis is based in the observation that those high on facets of prosocial orientation—such as empathy—tend to show highly embodied responses to care-related stimuli. For example, it was recently found that watching a hand being pricked with a needle triggered a highly embodied response— decreased motor excitability in the specific muscle targeted (i.e., the same motor response that occurs when one’s own hand is pricked with a needle; Avenanti, Minio-Paluello, Bufalari, & Aglioti, 2009). Critically, the magnitude of this embodied motor response correlated positively with trait empathy, a facet of prosocial orientation (Hoffman, 1977). Although this effect was indicative of simulation, it may also reflect a more general propensity for prosocially oriented individuals to have embodied responses of all kinds to care-relevant stimuli. Indeed, one of the major theories of empathy proposes that empathic responding is mediated by an embodied, perception-action link (Preston & de Waal, 2002). In a similar vein, we suggest that prosocially oriented individuals may possess a highly embodied concept of care such that care-related cues— such as cuteness—readily engage various bodily systems, including the motor system. If so, then these individuals may be especially likely to have cuteness trigger care-consistent changes in the motor system. For prosocially oriented individuals, the state of “care” that is elicited by cuteness may be more than a phenomenological and motivational state; it may also be an embodied state characterized by physically careful action. Research on care-related embodied responses has focused almost exclusively on responses to overt distress or harm (e.g., Singer et al., 2004). In this previous research, the affective care stimulus has been distress and the embodied response of interest has been simulation of that distress. But overt harm in others is not the only major care stimulus; a happy child, for example— on account of its “cuteness”— can elicit potent care-related impulses (Alley, 1983; Glocker et al., 2009; McDougal, 1908/1923). Thus, our current research extends previous research in terms of both the care stimulus studied (cuteness) and in the type of embodied response measured (physical carefulness). If our hypothesis is supported, it would suggest that the affective mechanisms underlying care may be even more extensively embodied than previously thought.

Measuring Embodied Care In the original demonstration of the cuteness-carefulness effect we used the game “Operation” as the measure of physical carefulness (Sherman et al., 2009). For this research, we developed a computerized analogue of this task: a tracing task that measures one’s ability to carefully manipulate a computer mouse to trace a path on the computer screen with the cursor. It is important to note that a recent study tested the effect of viewing positive and negative images on a similar tracing task (Coombes, Janelle, & Duley, 2005). This research revealed no consistent effects of emotional valence. In one study, participants made more errors after viewing unpleasant images than after viewing positive images. In a second study, however, there was no difference in performance across conditions. Most critically, there did not appear to be any consistent effects relative to neutral images.

Overview of Studies In two studies, we tested the hypothesis that prosocially oriented individuals—those with a predisposition to prioritize the welfare of others—would be especially likely to respond to cuteness with heightened physical carefulness. To test this hypothesis, we exposed participants to images of human infants—the cuteness of which we manipulated using photomorphing software—and then assessed performance on the cursor tracing task. Although the main performance measure was precision (deviation from the path), it is possible that physical carefulness could also manifest in terms of the time it takes participants to complete the task (i.e., duration). That is, becoming more physically careful may cause one to slow down. As result, we report the effects of cuteness on both precision and duration. To ensure generalizability, we measured prosocial orientation in two ways: (a) the endorsement of values that prioritize entities other than the self, including close others, humanity, and nature (Study 1); and (b) the tendency to consider harm a central consideration when judging the morality of acts (Study 2). If the predictions hold across these two different measures, this would strengthen our confidence that the effect is not particular to any single measure but rather is due to an underlying psychological trait common to both measures—the tendency to be concerned with the needs and suffering of others.

Study 1 In Study 1, we measured prosocial orientation using the Schwartz Values Survey (Schwartz, 1992), which assesses 10 values, clustered into subgroups. The care-related cluster—which Schwartz calls “self-transcendence”— consists of the values of benevolence (concern with close others) and universalism (concern with all of humanity and nature). These values predict various prosocial behaviors, including cooperation in an economic game (Schwartz, 1996) and charitable donations to a human rights organization (Verplanken & Holland, 2002).

Participants Participants were 311 (180 male, M age ⫽ 35.13) registered visitors to YourMorals.org—a research website offering various questionnaires and tasks related to morality—who completed both the Cursor Tracing Experiment and the Schwartz Values Survey.

Stimuli To create the stimuli for the studies, we first collected images of infants and toddlers (excluding newborns) from a stock photography website (the “portraits/children” section of http://www .sxc.hu). We then selected eight high-quality images to morph using Fantamorph photo-morphing software. We created the morecute images by making the eyes, cheeks, and forehead slightly larger, and the nose, lips, and chin slightly smaller. We created the less-cute images, by adjusting these features in the opposite direction. These features were selected because they have been shown to influence judgments of cuteness (Alley, 1981, 1983; Gross, 1997; McCabe, 1988). To validate the morphing manipulation, an independent sample (N ⫽ 756) rated the morphed images (and originals). We created

CUTENESS AND EMBODIED CARE

three different sets of photographs, each containing a mixture of originals, less-cute versions and more-cute versions (no set contained more than one version of the same image). Participants were randomly assigned to view one of the sets. They rated each child’s cuteness on a 7-point scale (1 ⫽ not at all, 7 ⫽ extremely) and each child’s age on a 10-point ordinal scale (1 ⫽ 3 months, 10 ⫽ 48 months) in which the first four options were in 3-month intervals (3, 6, 9, and 12 months) and the rest in 6-month intervals (18, 24, 30, 36, 42, and 48 months). As predicted, the more-cute version of each photograph was judged cuter and younger than the less-cute version. Compared with the average less-cute image, the average more-cute image was judged to be cuter (M ⫽ 4.96 vs. M ⫽ 4.28, p ⬍ .001, d ⫽ 0.49) and younger (M ⫽ 2.73 vs. M ⫽ 3.95, p ⬍ .001, d ⫽ 0.95).

Materials and Procedure The Schwartz Values Survey contains 57 values phrased as desired states or traits, which are rated for importance (8-point scale; ⫺1 ⫽ opposed to my values, 0 ⫽ not important, 7 ⫽ of supreme importance). We created a single prosocial orientation score for each participant by averaging the 15 self-transcendence items (␣ ⫽ .83), which include items describing states and traits focused on helping others and caring for the weak. To test for specificity, we also measured and tested the role of the other eight values, which were power (␣ ⫽ .77), achievement (␣ ⫽ .66), self-direction (␣ ⫽ .61), hedonism (␣ ⫽ .83), stimulation (␣ ⫽ .77), tradition (␣ ⫽ .63), conformity (␣ ⫽ .70), and security (␣ ⫽ .65). Whenever prosocial orientation emerged as a significant moderator of the effect of cuteness level (on either explicit cuteness ratings or physical carefulness) we also tested, separately, the other eight values as potential moderators. The study appeared on YourMorals.org under the title, “Cursor Trace: How well can you trace a line?” Participants first viewed eight photographs, which were presented in random order. Cuteness was manipulated between participants: depending on condition, participants saw either all of the less-cute images or all of the more-cute images. Duration of image presentation (4 or 8 s per image) was randomly assigned at the participant level and did not influence any results. Next, participants were told that their precision using the mouse cursor would be tested. For each trial, a black line 5 pixels wide was randomly generated creating a vertical path (M length ⫽ 1006.50 pixels, min ⫽ 408, max ⫽ 1558), which participants traced from top to bottom. Each line changed direction several times, always at right angles, thereby creating a path that had both vertical and horizontal segments. The program was able to detect the size of each participant’s browser window and generated a path that did not exceed these dimensions (i.e., the entire path was visible within the browser window; no scrolling was necessary). Participants were told, “Please trace as close to the black line as possible. Your score for each line will reflect how much you deviate from the center of the black line. There is no time limit.” To prevent skipping ahead, if participants deviated from the line by more than 10 pixels, the task momentarily stopped, a red box appeared at the point of departure from the line, and the following text appeared: “You have gone too far outside the line. Please return to the red box to return to the task.” As soon as the participant returned to the red box, the task resumed. Each participant traced five different lines. For each line, we recorded

153

the average deviation in pixels. We averaged these scores to create a final physical carefulness score (␣ ⫽ .82), which we then inverted and z scored so that higher scores indicated greater carefulness. We also created a similar score representing average duration (␣ ⫽ .75) per line (adjusted for length), with higher numbers indicating more time spent on the task. Afterward, participants indicated the type of computer mouse they had used. They were presented with five choices (each accompanied by a photograph): mechanical, trackpad, optical or laser, trackball, and other. The most common types were optical/laser (n ⫽ 176), trackpad (n ⫽ 88), and mechanical (n ⫽ 34). Finally, participants rated how cute and how interesting they found the photographs (9-point scale, 1 ⫽ not at all, 9 ⫽ extremely; one rating per dimension, rating the images “as a whole”) as well as their current mood (1 ⫽ extremely negative, 9 ⫽ extremely positive). Admittedly, given the web-based setting of the study, there were numerous features of the participant’s experience (e.g., screen resolution, cursor speed) over which we had little or no control. Critically, however, any error introduced by these factors was randomly distributed across the experimental conditions. This loss of power— due to increased noise in the data—was at least partially counteracted by our ability to collect large samples (over 1,500 total participants across the two studies).

Results and Discussion Explicit Ratings In a regression model predicting cuteness ratings from cuteness level, gender (⫺1 ⫽ female, 1 ⫽ male), prosocial orientation, and all interactions, two effects emerged: a significant effect of prosocial orientation, ␤ ⫽ .33, p ⬍ .001, and a marginally significant effect of cuteness level, ␤ ⫽ .11, p ⫽ .07. This effect of cuteness level was not moderated by gender or prosocial orientation (ts ⬍ 1). Thus, being female or high on prosocial orientation did not make one more sensitive to subtle differences in cuteness. Cuteness level did not influence how interesting participants deemed the images (t ⬍ 1.19) but did influence mood. Those who saw the more-cute images reported slightly better mood than those who saw the less-cute images (M ⫽ 6.02, SD ⫽ 1.48 vs. M ⫽ 5.65, SD ⫽ 1.57), t(307) ⫽ 2.13, p ⬍ .05, d ⫽ 0.24.

Physical Carefulness In a regression model that predicted physical carefulness from cuteness level, gender, prosocial orientation, and all interactions, there was only one significant effect: a Cuteness Level ⫻ Prosocial Orientation ⫻ Gender interaction, ␤ ⫽ ⫺.12, p ⬍ .05. This three-way interaction was specific to prosocial orientation. This was evident when computing a series of parallel models, each testing one of the other eight values from the Schwartz Values Survey. In none of these alternative models did the critical threeway interaction emerge (Fs ⬍ 2.72, ps ⬎ .10). We decomposed the significant three-way interaction by computing separate models by gender. There were no effects for men (ts ⬍ 1). For women, there were no main effects of cuteness level (␤ ⫽ .11, p ⫽ .24) or prosocial orientation (␤ ⫽ ⫺.07, p ⫽ .47), but there was a significant Cuteness Level ⫻ Prosocial Orientation interaction, ␤ ⫽ .20, p ⬍ .05. This interaction remained significant

SHERMAN, HAIDT, IYER, AND COAN

154

with duration as a covariate. According to simple slopes analysis (Aiken & West, 1991), viewing the more-cute images enhanced physical carefulness for women high on prosocial orientation (1 SD above the mean), ␤ ⫽ .34, p ⬍ .01, but had no effect on women low on prosocial orientation (1 SD below the mean), ␤ ⫽ ⫺.05, p ⫽ .67. This pattern of results is summarized in Table 1. Analysis of participants’ mood ratings suggests that general positivity is unlikely to explain the effect of cuteness on physical carefulness. The three-way interaction (Cuteness Level ⫻ Prosocial Orientation ⫻ Gender) remained statistically significant (␤ ⫽ ⫺.14, p ⬍ .05) with mood (and all interactions with mood) entered into the regression model. In addition, the simple effect of cuteness for prosocially oriented women remained statistically significant (␤ ⫽ .34, p ⬍ .01) when controlling for mood, suggesting that the mood-enhancing effect of the highly cute images cannot explain their impact on physical carefulness.

Duration To test whether carefulness was also apparent in the time it took participants to complete the task, we repeated the analysis with duration as the outcome. This analysis revealed a marginally significant main effect of prosocial orientation (␤ ⫽ .11, p ⫽ .06; greater prosocial orientation associated with taking longer to complete the task), a main effect of gender (␤ ⫽ ⫺.20, p ⬍ .001, women spent more time on the task than men), and the three-way Cuteness Level ⫻ Prosocial Orientation ⫻ Gender interaction (␤ ⫽ ⫺.14, p ⬍ .05). To test whether duration showed the same pattern as the precision measure, we decomposed the interaction, computing separate models by gender. For men, there was a main effect of prosocial orientation (...