Eyes as windows to the soul: Gazing behavior is related to personality

Abstract

Gazing is a fundamental human behavior with important cognitive, affective, motivational, and social underpinnings that is likely to have produced individual differences linking it to major personality traits. If traits play a substantial role in gazing, they should predict eye movement parameters above and beyond stimuli without meaningful and topical information. The current eye-tracking study (N = 242) demonstrated with linear mixed models that personality (Big Five, Behavioral Inhibition System/Behavioral Activation System) predicts number of fixations, mean fixation duration, and dwelling time in two different abstract animations. Specifically, neuroticism, extraversion, openness, and the Behavioral Activation System were related to eye movement parameters. Prospective research in studying links between dispositions and gazing is discussed.

Introduction

According to a popular proverb, “eyes are the window of/to the soul.”² And, indeed, people have long pondered whether there is something in our eyes indicative of character. For example, someone often socializing would naturally be called “extraverted,” but could we also deduce extraversion from how someone gazes? Behaviors are core features of traits, personality manifests at different levels of behavior, and gazing is an essential human behavior. The current work examines associations among traits and eye movements to gain new insights on whether and how dispositional variables manifest in oculomotor behavior.

Gazing behavior can (and should) be linked to personality for several reasons. First, individual differences manifest on molar and molecular levels of behavior (Furr, 2009). Fleeson and Noftle (2008a) point out that there is only “very little knowledge about how personality is present in behavior and about what behaviors are relevant to personality,” which is “partly because of the difficulty in specifying the level at which behavior should be studied” (pp. 1668/1679). This problem in identifying personality-relevant behaviors is also crucial for other psychological disciplines such as social and experimental psychology which tend to consider possibly important individual differences only to a minor degree. Attention should also be given to individual differences in micro-behaviors such as eye movements.

Second, gazing serves important social, motivational, cognitive, and regulatory functions. Eye contact is essential in our daily lives: it is related to personality (Libby & Yaklevich, 1973) and we make inferences about people based on their eye contact (Brooks, Church, & Fraser, 1986). Gazing is used to track others’ behaviors (Matsumoto, Shibata, Seiji, Mori, & Shioe, 2010) and to communicate. It is also linked to motives and motivation (e.g., Furtner et al., 2011, Terburg et al., 2011), wishes, and preferences as it is directed toward goal-consistent and averted from goal-inconsistent stimuli to achieve and maintain good mood (Balcetis and Dunning, 2006, Isaacowitz, 2006). Eye movements can also be used for perceptual-cognitive performance tasks (e.g., Haley, 1971, Iacono and Lykken, 1979) because they are indicative of early visual attention processes guided by preconscious mechanisms (Terburg et al., 2011; Wilkowski, Robinson, Gordon, & Troop-Gordon, 2007). Moreover, precise eye-hand coordination is an important feat in humans (Furtner & Sachse, 2008).

Third, stimulus material substantially influences eye movement parameters such as fixation rates (Land, 2007), but participants nonetheless tend to show stable eye movement patterns (across different stimuli/occasions) and differ in those from others (Etaugh, 1972, Etaugh and Rose, 1973, Furtner, 2006). This implies an underlying neurobiological system influencing oculomotor (re-)activity, which is linked to personality (Canli, 2006).

Fourth, eye movements have already been linked to personality. Kaspar and König (2011) found that interindividual differences in personality and motivation influence attention processes in gazing, and personality psychopathology has also been related to eye movements (Ceballos and Bauer, 2004, Iacono and Lykken, 1979, Siever, 1982, Siever et al., 1990, Thaker et al., 1996). Also, lateral eye movements (averting the gaze while thinking; Day, 1970, Etaugh, 1972, Etaugh and Rose, 1973) and attentional preferences in selective attention during early visual processing have been linked to circumscribed personality dimensions (optimism: Isaacowitz, 2005; trait anger: Wilkowski et al., 2007).

In summary, previous research suggests that personality influences visual information processing, social gazing, and where we gaze when making sense of (ambiguous) pictures. However, there are limitations to this research. First, not all studies were able to use sophisticated eye-tracking methodology, which provides objective and quantifiable behavioral data. Second, rather small sample sizes were used which present difficulties in detecting small effects. Effect sizes in trait – eye movement relations ought to be in the range of r = |.05 − .30|, which can be deemed common and realistic when broad, self-rated traits are related to narrowly defined, objective behavioral data on a micro-level (Vazire & Carlson, 2010). To demonstrate such effects, usually larger sample sizes (N > 150) are needed. Third, many studies do not focus on sub-clinical traits, and if they do, only on very specific ones (cf. Isaacowitz, 2005, Terburg et al., 2011, Wilkowski et al., 2007) which does not allow a bigger picture on trait – eye movement relations. Lastly and most importantly, it remains unexplored whether dispositional variables are also related to the “how” of gazing (e.g., more fixations), not just the “where” (e.g., fixating a car). Personality effects in the absence of meaningful stimuli with semantic or topical information would provide compelling evidence that traits are linked to gazing. The current study is a first endeavor in demonstrating such effects.

The broad Big Five traits (neuroticism, extraversion/surgency, openness/intellect/fantasy/culture, agreeableness, conscientiousness) were used for several reasons. First, the structural and descriptive five factor model of personality is a widely acknowledged, integrative taxonomy of most human individual differences categories that are important, meaningful, and consequential (Costa and McCrae, 1992, John and Srivastava, 1999).

Second, the Big Five have a biological basis (Angleitner and Ostendorf, 1994, Buss and Plomin, 1975, DeYoung and Gray, 2009). Indeed, each of Cloninger’s temperament and character factors³ (Cloninger, Svrakic, & Przybeck, 1993) is substantially covered by the Big Five traits due to considerable construct overlap (de Fruyt, van de Wiele, & van Heeringen, 2000). Extraversion (linked to positive affect and activation) and Neuroticism (linked to negative affect and affective intensity) are considered strongly biologically determined (MacDonald, 1995, Rothbart and Derryberry, 1981, Yik and Russell, 2001). Agreeableness and conscientiousness may be traced back to effortful control, a super-ordinate regulatory system (Jensen-Campbell et al., 2002). The biological basis and associations with affect and activation may link the Big Five traits to eye movements.

Third, there is already empirical evidence that some of the Big Five can be linked to gazing. Openness has been linked to eye fixation points (Matsumoto et al., 2010), and extraversion to performance on anti-saccade tasks (Nguyen, Mattingley, & Abel, 2008) and spontaneous eye movements such as blinking rates (Franks, 1963). However, no study so far has linked the entire Big Five traits at once to eye movements.

The Behavioral Inhibition System (BIS), regulating sensitivity to punishment and avoidance behavior, and the Behavioral Approach/Activation System (BAS), regulating sensitivity to reward and approach behavior (Carver and White, 1994, Corr, 2008, Gray, 1987, Gray, 1990, Gray, 1991), can also be tied to eye movements. The BIS/BAS is rooted in Gray, 1982, Gray, 1991, which has been revised in the meanwhile. The BIS can be seen as a basis for anxiety (Gray & McNaughton, 2000) and the BAS for reward sensitivity/approach and impulsivity (Gray, 1991). First, BIS/BAS are considered neurobiological underpinnings of affective and motivational systems (Cloninger et al., 1993, Gray, 1991). Second, BIS/BAS can be linked to psychopathology (Scholten, van Honk, Aleman, & Kahn, 2006), which is known to influence oculomotor behavior (e.g., Iacono & Lykken, 1979). Third, BIS and BAS can be linked to the Big Five: consistent findings for relationships with neuroticism (for BIS) and extraversion (for BAS) are found (Keiser and Ross, 2011, Smits and Boeck, 2006). If the Big Five are associated with eye movements, then BIS and BAS should also be associated given their conceptualization in literature and empirical relations with the Big Five.

Because not all readers may be familiar with eye-tracking methodology, we provide brief information on eye movement analyses. For more detailed introductions, see, for example, Duchowsky (2007). Visual perception relies on sequences of information input via complex patterns of eye, head, and body movements (Furtner & Sachse, 2008). High-resolution visual information input occurs only at so-called fixation points which are fixated with the fovea (the central point of highest visual resolution in the retina) (Posner, 1980, Posner, 1995). Saccades lie between fixations and entail a passive suppression of visual information processing during which visual perception is strongly reduced for about 50–80 ms. Eye movements can be recorded and analyzed with two basic types of eye-trackers: remote or table-mounted and head-mounted systems. Remote systems are usually affixed to a table with video cameras and infrared lights (Furtner et al., 2009, Goldberg and Wichansky, 2003, Jacob and Karn, 2003), whereas head-mounted systems are worn on the head (Duchowsky, 2007). Data recording is usually carried out with the pupil center corneal reflection method (Ohno, Mukawa, & Yoshikawa, 2002) where the eyes’ position and direction of movement is related to a vector (visual axis) spanning from the corneal reflection (Purkinje reflection), captured with infrared light, to the center of the pupil. There are many different eye movement parameters (e.g., saccade duration, saccadic velocity, saccadic acceleration, saccadic amplitude, smooth pursuits, fixation duration, number of fixations, gaze duration, dwelling time; see Joos, Rötting, & Velichkovsky, 2003), but a meta-analysis has shown that the three most widely used eye movement parameters are number of fixations, mean fixation duration, and dwelling time (Jacob & Karn, 2003; see also Joos et al., 2003), which all refer to fixations and can be sampled with the pupil center corneal reflection method (Ohno & Mukawa, 2004). In the current study, we also utilize a table-mounted eye-tracker with pupil center corneal reflection method to capture eye movement parameters.