Realness is a core feature of authenticity

We established realness as the relatively stable tendency to act on the outside the way one feels on the inside, without regard for proximal personal or social consequences. In nine studies, we showed that real-ness is a) a core feature of individual differences in authenticity, b) generally adaptive but largely unrelated to agreeableness, c) highly stable, d) reliably observable in dyadic behavior, and e) predictive of responses to situations with potential for personal or social costs. Informants both perceive agreeable motives in real behavior and recognize that being real can be disagreeable. We concluded that realness represents an important individual difference construct that is foundational for authentic social behavior, and that being real comes with both costs and beneﬁts.


Introduction
There are times in life when it is difficult to know whether or not to reveal your true thoughts, feelings, and desires.For instance, what do you do when you are annoyed by a friend who is being rude to restaurant staff?Or, what do you do when a new romantic partner is publicly affectionate in a way that makes you feel uncomfortable?How do you handle a situation in which someone you know is clearly upset by something, but seems reticent to tell you what the problem is?What if you like someone but are too nervous to say so, and find yourself instead expressing your liking by teasing them, and giving the wrong impression?Being explicit in these situations comes with risks for yourself and others.It can make situations uncomfortable or awkward, or convey the wrong impression.This can, in turn, have negative long-term effects.On the other hand, it can feel fake to hide the way you think or feel, and this can also come with both short-and long-term costs.Most people can readily identify friends who would almost certainly confront a rude friend, establish boundaries about public affection with a new date, ask someone what is bothering them, or express genuine liking in situations like those described above.For such people, expressing themselves directly seems to trump other concerns.These kinds of people are real, even in moments with potential personal and social costs.
We use the term realness to mean behaving on the outside the way one feels on the inside, without regard for proximal personal or social consequences.We situate realness within the broader concept of authenticity.However, whereas we understand authenticity as a dynamic, multidimensional process, we conceptualize realness as a more specific, core feature of that process that is revealed only in certain social conditions.Specifically, we understand realness as important for authenticity because it reflects stable individual differences in the tendency to be authentic when situations apply pressure to do otherwise.This pattern of behavior and its potential downsides was a major emphasis of classical models of authenticity that is mostly missing in contemporary empirical research and assessment tools.In particular, realness has been hidden by a strong positive valence in authenticity questionnaires and efforts to include other features that may support authentic behavior, such as inner values, self-awareness, or various styles of external expression.
In the introduction that follows we describe how realness has been described in classical theories of personality, with particular attention to distinguishing realness from the broader and more complex construct of authenticity.We then review empirical work on authenticity, to highlight how realness has become obscured by methodological and conceptual issues in contemporary research.We then present a series of nine studies in which we generate a measure of realness by uncovering the core and common behavioral dimension within existing authenticity measures, removing the positive valence from that dimension, and examining how it relates to a variety of theoretically relevant variables.We conclude based on the results of these nine studies that realness represents an important individual difference construct that is foundational for authentic social behavior, and that being real comes with both costs and benefits.

Realness in personality theory
A review of classical theories of realness and related concepts in the personality literature reveals certain trends and consistencies regarding its nature and consequences.

Realness is healthy
Realness has been regarded as a principal outcome of healthy development and/or effective psychotherapy, a point that has been especially stressed by clinically oriented theorists.Jung (1939) described a true self that lies beneath defensive personas and that is capable of growth and contact with meaningful symbolism.Winnicott (1958) distinguished the true vs. false self in personality, the former being whole and unified whereas and the latter at risk of splitting, dissociation, and incoherence.Horney (1951) asserted that the real self is ''the alive, unique, personal center of ourselves, the only part that can and wants to grow" (p.155) but warned that ''under stress, the person will become alienated from his real self '' (p.13) whereas ''if nourished, the real self surges toward selfrealization" (p.17).Early existentialists held this pattern of behavior up as among the highest touchstones of moral achievement (Boss, 1963;Heidegger, 1927;Kierkegaard, 1849), with outcomes that not only reflect personal health and well-being, but which also have positive effects for relationships and society.These perspectives highlight that realness has generally been considered, by theorists of various persuasions, as an outcome of psychological maturation, health, and adjustment.

Realness has downsides
Theorists have also consistently stressed the balance between the upsides and downsides of authentic behavior (see Jongman-Sereno & Leary, 2019).There is a reason people often censor what they say or how they behave; unmitigated realness can hurt or embarrass other people or the self.In personal relationships, social tact often involves holding back, whereas people often regret having revealed how they truly feel during moments of emotional dysregulation.At a social level, being real in the political sphere can garner support from a politician's base but sow divisions at the national level (Rosenblum, Schroeder, & Gino, 2019).
The personal downsides of benevolent realness have received significant theoretical attention.For Kierkegaard (1844), being authentic inevitably conflicts with being a reliable member of social institutions.It is lonely, alienating, and produces feelings of dread.Rogers (1961) put a fine point on this aspect of authenticity by emphasizing that it requires lessening the influences of ''oughts", expectations, and needs to please others.He urged people in the direction of autonomy despite various pressures to fit in.Maslow (1968), who thought of authenticity as the cardinal behavioral indicator of self-actualization and viewed the authentic person as ''complete and final in some sense" (p.123), asserted that ''he" (sic.) must ''transcend his culture" (p.16) and generally underscored that risks of self-actualization include social ostracism and being seen as proud, arrogant, or indifferent.He also warned how self-actualized realness may not be particularly popular, even if the self-actualized person is working towards social causes larger than themselves (the upshot of his perspective, though, is that the self-actualizing person will not care all that much).May (1953, p. 193) asserted that ''people lack courage (to be their authentic selves) because of fear of being isolated. ... laughed at, ridiculed, or rejected" and held up Socrates and Spinoza, two philosophers celebrated for rejecting social norms at tremendous personal costs in order to achieve authentic expression, as prototypes.
Indeed, many famous martyrs for noble causes suffered greatly from being real.Martin Luther King, Jr. was murdered by a reactionary and Galileo was subjected to an undignified funeral as a consequence of their efforts to hasten human progress by speaking truth to power.Sojourner Truth and Thomas Paine advanced their causes with unusual risk and valor and at tremendous personal cost.The #MeToo Movement or American athletes who kneel during the U.S. national anthem provide more contemporary examples that are surely complemented by less public instances that the reader could easily call to mind.Of course, there are also many converse examples, in which people with antisocial motives caused harm by being real.
In contrast, there may also be advantages to certain forms of inauthentic behavior.Deceit has been taken for granted and even extolled as a political mechanism (Machiavelli, 1513(Machiavelli, /1916;;von Rochau, 1853).A particularly nefarious version of deceit has to do with seeming to be real to one's political base, in a way that is offensive or divisive to society in general.For instance, populist leaders have summoned racist or classist instincts among citizens as a tactic to engender support, with the implication that they are simply being honest rather than politically correct.To be clear, we do not know if such individuals were being real (whether they were personally racist or classist) or whether they were using a politically advantageous strategy, just as we cannot know if people who are being agreeable are being real when it is to their advantage.But in many cases, there is reason to suspect that the strategy was more important in such people than an honest expression of their inner values (we leave room for the likelihood that some world leaders have simply been really bad people).Humor, which is often literally or at least concretely inauthentic (e.g., in sarcasm and irony a person says things they do not mean) is generally associated with positive outcomes (Kuiper, Martin, & Olinger, 1993), particularly when it is good natured (Barnett & Deutsch, 2016;Leist & Müller, 2013).Lerner (1993) suggested that it may be adaptive for women trying to cope with sexism to avoid realness, at least in certain contexts, because asserting one's views directly can have disparately negative attributions for women relative to men.These examples augur a second commonality in classical theories of realness: it may have both good and bad consequences for the individual and society.In particular, in some cases realness can come across as disagreeable or impolite, disrupt social harmony, and alienate the person who has been real, even if a person has good intentions.Conversely, if a person truly has unseemly thoughts and feelings, expressing those thoughts and feelings, or being real, can cause harm to self and others, particularly to the degree that such individuals are politically powerful.

Being real is a core aspect of the broader concept of authenticity
The third thread in this tapestry is more complicated still.What are the processes and components that give rise to behavioral realness, and how do they fit together?From an existential perspective, the essence of personal development involves escaping the nihilism that comes with recognizing that there is no objective purpose to life, and creating a subjective meaning to which personal energy, values, and behaviors can be attached (Sartre, 1946).Fundamental to this essence is the use of one's inner life to guide external behavior.Rogers famously described this construct -which he variously referred to as genuineness, congruence, authenticity, and realness -as both a principal outcome of effective therapy (Rogers, 1961, p. 165) and one of the three necessary and sufficient characteristics of effective psychotherapists (Rogers, 1957).He asserted that realness occurs when ''the feelings the person is experiencing are available to him (sic.),available to his awareness, and he is able to live these feelings, be them, and is able to communicate them" (Rogers, 1961p. 61).From his perspective, being real allows a person to accept whatever comes their way and act in a way that is adaptive, because ''he" has ''trusted his gut" and acted upon his inner experience in a specifically specified class of situations.
The view that realness has essentially to do with acting on the outside the way one feels on the inside is common (c.f., ''genuineness" in Ryan & Ryan, 2019) but it has typically been wrapped in the trappings of a more complicated, dynamic, multifaceted process, under the rubric of authenticity.For instance, the central distinction of Deci and Ryan's (1985;2000) highly generative Self-Determination Theory (SDT) is between authentic and external motivations for behavior.In the SDT model, authentic motivations are intrinsic and self-authored goals organized to achieve a sense of competence, relatedness, and autonomy.Authentic motives are at perpetual risk of stultification by external, environmental motives; seemingly benign or even positive reinforcers like verbal praise, financial compensation, or public reward can dull inner motives for authentic living (Deci, Koestner, & Ryan, 1999).Kernis and Goldman (2006) understand the disposition to be authentic as involving four components that are measured using separate scales on their questionnaire: being aware of one's inner states, processing those states in a way that is relatively free from bias, expressing those states in behavior, and being particularly motivated to be genuine in close relationships.Wood et al. (2008) conceptualize authenticity as a process of a) an inner experience of self, b) accurate awareness of that inner experience, and c) expression of that inner experience.Their questionnaire contains three corresponding scales: self-alienation, authentic living, and accepting external influences.The idea is that self-alienation prevents accurate awareness, and that either variation in the motive to be authentic (authentic living) or susceptibility to other motives (external influences) can interfere with authentic expression.Sheldon, Davidson, and Pollard (2004) apprehend authenticity as a character trait that describes people who are true to themselves and accurately represent their internal states (feelings), intentions (thoughts), and commitments (behaviors).Although there is variation in these sub-scale structures, they have in common a distinction between internal (i.e., awareness) and external (i.e., behavior) domains of authenticity (Knoll, Meyer, Kroemer, & Schröder-Abé, 2015).
Overall, these models share the assumption that authentic behavior is the result of a dynamic, multicomponent process.Moreover, even though they may not have common views about what those components are or their relative importance, they all emphasize the connection between inner and outer states.In other words, these models disagree about which specific internal and external features are contained within the authenticity construct, whereas they agree that the connection between these features is critical.We concluded from this literature that authenticity, as a complicated, multi-component, temporal, and highly contextualized process, is unlikely to be captured in cross-sectional questionnaire data.However, the connection between the way people feel on the inside and how they behave on the outside, independent of the internal or external components themselves, represents a core aspect of authenticity.The premise of this paper is that isolating this aspect would be a valuable step toward a more empirically tenable model of authenticity that is faithful to classical theories of authenticity.

Summary
Based on our literature review, we define being real as behaving on the outside the way one feels on the inside, without regard for personal or social consequences.It is distinguishable from other features of authenticity and other personality variables by several properties.First, authenticity is a complex construct with internal/psychological and external/behavioral dimensions (Fig. 1).Internal aspects include psychological functions that support authentic behavior, such as self-awareness, accuracy of social perception, and capacity for reflection.External behavior includes all of the verbal and non-verbal expressions that communicate variation in authenticity to others in social situations.Realness is the connection between these internal and external dimensions.When people act the way they think and feel (whether those internal states are positive or negative, conflictual or straightforward, socially acceptable or not), they are being real.Second, realness is a product of psychological maturation, and thus should be positively associated with indicators of wellbeing, mental health, and mutually satisfying relationships.This is particularly the case among people who have relatively developed internal functions as depicted in Fig. 1.Third, however, realness may involve violating social norms, and thus it can come with both upsides and downsides for self and others.In particular, realness reflects prioritizing being genuine over social harmony, and thus should not be positively associated with traits related to the personality domain agreeableness.On the other hand, one can also be fake by acting less agreeably than they actually feel, and many people appreciate realness in others.For these reasons, we would not expect realness to be negatively associated with agreeableness, either.Fourth, to the degree that realness can be used to describe people, in general, realness scores should be relatively stable over time, albeit subject to the influence of situational and developmental factors (similar to personality traits; see Bleidorn & Schwaba, 2017).Fifth, given that it manifests in social contexts (Chen, 2019), realness should be observable by others, meaning that other people should be able to reliably rank people they know in terms of their tendency to be real.Sixth, individual differences in realness should predict behavior in situations when there are plausible negative consequences for being real, because realness is not simply being disagreeable and simply upsetting people, nor is it being honest when it is to one's advantage, it is a pattern of behavior that proves itself by being, at times, socially disadvantageous.

Realness within empirical research on authenticity
Given its theoretical importance for social behavior, it is not surprising that there has been both longstanding interest in authenticity among personality and social psychologists and a recent swell of research on the subject (Kovács, 2019;Hicks et al., 2019;Hutchinson, 1995).We review that work in this section in terms of the properties summarized above, to examine the degree to which realness as we conceptualize it is represented therein.
Existing authenticity measures can be categorized into two types: those that measure a general disposition to be authentic (Kernis & Goldman, 2006;Knoll et al., 2015;Wood et al., 2008) and those that measure the tendency to be authentic in particular roles or relationships (Brunell et al., 2010;Fleeson & Wilt, 2010;Gelso, 2002;Sheldon, Ryan, Rawsthorne, & Ilardi;1997).When these measures have multiple dimensions, their scales generally distinguish between more internal and more external features of the construct, as reflected in Fig. 1.As discussed above, these components are thought to interact in a dynamic, temporal process.Although models differ with regard to specifics, the general idea is that the person becomes aware of some inner goal, sense, or experience then acts upon that awareness in a manner that is more or less true to the inner goal, sense or experience.
We have two concerns about using these kinds of tools to approximate this sort of process.First, we question the centrality of internal and external features, in isolation, to the concept of authenticity.Internal features, such as having some level of awareness of internal states, being able to accurately perceive external contexts, reflection, and self-regulation undergird a wide variety of adaptive psychological functions.One of these functions may be a tendency toward authentic behavior, but authenticity is neither a unique nor a necessary outcome of these healthy internal processes.For instance, we would expect relationship success, well-being, and lower risk for psychopathology to also follow from these capacities.Although external behavior is a critical indicator of authenticity, it also is not sufficient without reference to internal states.We thus focus on the connection between inner states and outer behavior, as discussed above.
Second, we are skeptical that it is possible to decompose a complex dynamic within-person process using measures that assess relatively stable attributes designed to make between-person distinctions at one point in time.It is perhaps telling that, in contrast to the multidimensional questionnaires that are typical in crosssectional authenticity research, studies focused on the experience of authenticity in particular moments (e.g., Fleeson & Wilt, 2010;Lenton, Bruder, Slabu, & Sedikides, 2012;Sheldon, Gunz, & Schachtman, 2012) or observations of others' authenticity (Gershon & Smith, 2019;Rosenblum et al., 2020) tend to gauge authentic experience as a single construct that implicitly amalgamates inner goals, senses, experiences, awareness, and behavior.That is, whereas research that does not attempt to capture temporal processes makes distinctions about internal and external aspects of authenticity that are presumed to dynamically interact, research that endeavors to study authenticity as a function of situational context tends to collapse different components.Thus, neither cross-sectional nor temporally sensitive research has fully accounted for the processes theorized to account for authentic personality and interaction styles in terms of both dynamic processes and distinctions between awareness and behavior.
To be clear, we largely accept the thesis that different psychological components likely interact with one another in a variety of complicated ways that ultimately give rise to what is regarded as authentic patterns of behavior.However, our position is that the research literature may have gotten ahead of itself by trying to capture this entire process with multidimensional questionnaires, typically implemented in cross-sectional designs.We concluded that it would be better to build up from robust, unitary and foundational concepts to more complex, temporally nuanced models.This would enable a foundational literature on the basic parameters of this essentially authentic tendency.Moreover, establishing the structure of different elements of authenticity realness should support eventual work designed to unpack how these elements unfold within dynamic and highly contextualized social processes.As such, we focus on realness as the core dimension of authenticity.While we accept the general notion that authenticity involves other constructs, we do not have a position about the degree to which existing models accurately capture the constructs involved.Instead, we focus on isolating realness as a core and essential feature of authentic behavior.

Hypothesis 1. Realness content will be present in existing authenticity measures
Given that we see realness as a core distinguishing feature of authenticity, we should expect to find realness item content within existing authenticity measures, albeit perhaps obscured and in need of adaptation to clearly demarcate the tendency to act on the outside the way one feels on the inside, without regard for personal or social consequences.
Hypothesis 2. Indicators exclusively focused on realness will form a unidimensional trait

Correlates of authenticity
Although realness should be reflected in thoughts, feelings, desires/goals, and behaviors, this core should cohere as a single, homogeneous construct.Realness can manifest through the congruence between inner emotions and outer affects, the transparency of one's motives, the sharing of perceptions and attributions, or directly through behavior as perceived by others.It is thus important that a robust model of realness include content related to each of these psychological functions.That being said, distinguishing these functions as separate factors was not our goal.Rather, our focus was on what they have in common.
Hypothesis 3. Realness should generally correlate with measures of well-being and adaptive functioning.
Research also suggests positive consequences over the longer term.From the perspective of SDT, authentic behaviors help to satisfy one's basic needs (Deci & Ryan, 1985;Ryan & Deci, 2000).Work by Sheldon and colleagues verifies that working towards more authentic goals is associated with well-being and goal attainment (Sheldon & Elliot, 1999;Sheldon & Houser-Marko, 2001;Sheldon et al., 1997), and a study by White and Tracey (2011) suggests that authenticity is related to confidence in career choices.Psychotherapy research connects patient ratings of their sense that the therapist is being genuine with treatment outcomes (Eugster & Wampold, 1996;Gelso, 2009;Marmarosh et al., 2009) perhaps in part because therapist self-disclosure helps clients see their therapists as more human (Knox et al., 1997).Overall, existing research leaves little doubt that authenticity is a net positive characteristic for well-being, social functioning, and adaptation.Given this pattern of authenticity correlates, the fact that we hypothesize that realness is a core feature of authenticity, and theoretical work positing realness as an outcome of healthy development, we expect realness to be generally adaptive as well.
Hypothesis 4. Realness should not be related to agreeableness traits.

Stability
That being said, it is also intuitive that most people think that it is better to be authentic than otherwise, and empirical correlates of various authenticity measures bear this out (Sedikides & Skowronski, 1997).Who would want to be seen, or to see themselves, as inauthentic (Jongman-Sereno & Leary, 2016;Lenton et al., 2012;2013)?Cross-cultural research suggests that authenticity is a relatively universal value (Slabu, Lenton, Sedikides, & Bruder, 2014).Like many positively-valenced variables, there is a risk for participants to see authenticity as an entirely positive attribute, as would be suggested by trait correlations that are all in the more adaptive direction (Leising et al., 2020).For instance, Jongman-Sereno and Leary (2016) showed in two studies that people view their positive actions as being more authentic than their negative actions, even when the objective authenticity of their behavior was controlled.
Critically, these correlations contrast with some of the nuance in influential theories of authentic personality reviewed above (e.g., Maslow, 1968;Rogers, 1961), suggesting that contemporary research may be missing something important (Baumeister, 2019).Specifically, in contrast to nearly universal positive correlations between authenticity measures and adaptive outcomes in empirical research, there is good reason to believe that there should be downsides, at least occasionally, to authentic behavior.In particular, early theorists would not support a definition of authenticity as being particularly agreeable, because there are times when being true to one's self risks disappointing, annoying, or frustrating others.Yet, Pinto et al. (2012) found that agreeableness was correlated between 0.23 and 0.49 with authenticity scales; these results included authenticity scales explicitly designed to measure more external or behavioral features.In our conception, realness should be reliably related to traits involving psychological adjustment (high conscientiousness and low neuroticism) and sociability (i.e., high extraversion and openness), but not to those related to a desire for social harmony and politeness vs. rudeness and antagonism (i.e., agreeableness).We note that this does not necessarily mean that observers will not see others' authenticity as warm or agreeable.Indeed, based on past research using informant-report methods, we would expect observers to prefer their friends to be real rather than artificially polite or superficially agreeable (e.g., Kovács, 2019;Liu & Perrewe, 2006).
Hypothesis 5. Individual differences in realness should be rank-order and mean-level stable across time at levels similar to personality traits.

Observability
Research suggests that authenticity measures achieve rankorder and mean-level stabilities that are in the range of what would be expected of individual difference constructs such as personality traits (Boyraz, Waits, & Felix, 2014;Reinecke & Trepte, 2014;Zhang, Zhou, Dik, & You, 2019).There is also evidence that people are more authentic in certain kinds of relationships (Robinson et al., 2018) and situations (Fleeson & Wilt, 2010;Lenton, Bruder, Slabu, & Sedikides, 2012;Sheldon, Gunz, & Schachtman, 2012) and that it increases with effective psychotherapy (Bayliss-Conway et al., 2020).Sheldon et al. (1997) found that the consistency of traits across different social roles and the degree of authenticity in these roles independently predicted overall wellbeing, suggesting that consistency in authentic behavior across situations may, itself, indicate an important characteristic of authenticity.Indeed, some authors have operationalized authenticity in terms of consistency across situations (Sutton, 2018).
This pattern of sensitivity to situational factors that is similar across individuals also accord with findings about personality traits (Fleeson, 2001).The overall pattern fits with the model shown in Fig. 1: realness, like other aspects of authenticity and personality traits more generally (Bleidorn et al., 2020), is a stable individual difference that is also responsive to situational and contextual factors.We thus expected individual differences in realness to be relatively stable across time, in terms of both rankings between people (r > 0.50) and absolute group changes (approaching 0) in the absence of external pressures.Hypothesis 6.Individual differences in realness should be observable in social situations.

Predicting situations
Given that authentic behavior occurs in a social context, realness should be observable by others, and observers should be able to reliably rank people as more or less real relative to one another.Several studies have demonstrated that observers can reliably code the authenticity in other peoples' behavior.For instance, Anderson et al. (2020) had people describe their emotional responses to a film clip they had just watched, and their descriptions were coded by three people using a single 7-point item ranging from genuine to not genuine.These codes were reliably related to one another (alpha = 0.67 and 0.74 in two studies).Gershon and Smith (2020) had participants rate entrepreneurs, politicians, tour guides, and comedians whose vignettes they read using three items asking about authenticity, sincerity, and genuineness (adapted from Barasch et al., 2014).The alphas for these items tended to be around 0.91 across several studies.Importantly, these authors also found that it was important to give observers access to multiple instances of targets' authentic behavior, because people actually rated targets as less authentic the more the targets repeated themselves.Rosenblum et al. (2019) found that observers reliably rated politicians who use politically incorrect language as more authentic (using adapted items from Wood et al., 2008 scale having to do with ease of being influenced) but also less warm.This finding is consistent with our view of realness in suggesting that assessments that focus more explicitly on observable behavior in a context with potential costs may reduce associations between authenticity and traits related to agreeableness.
Hypothesis 7. Individual differences in realness should predict behavior in situations where there are potential downsides to being real It is established that the state experiences of traits are reliably related to stable trait ratings (Fleeson, 2001), and a reliable correlation has been observed between ratings of trait authenticity and the frequency of experiencing authentic states (Lenton et al., 2012;Lenton, Slabu, & Sedikides, 2016).We note that a separate literature has to do with the fact that the felt experience of authenticity is more related to aspects of situations than consistency between one's behavior and one's self-rated personality traits (Fleeson & Wilt, 2010).This idea, while important for thinking about how authenticity manifests as a contextualized social process, is independent of our interest in identifying a marker of stable individual differences in realness.Taken together, existing research suggests that assessments of stable tendencies predict the likelihood of authentic behavior in specific social situations, on average.However, as noted above, realness can only be tested in situations in which there are potential personal or social downsides.When one's inner sensibilities and external pressures align, the same behavior can serve both goals.Thus, we should expect the finding that trait authenticity predicts situational behavior to extend to a more specific assessments of realness in situations with potential personal or social costs.

Overview of studies
We tested these hypotheses through a progression of nine studies (Table 1).In the first, we examined the subscales of three widely used multidimensional authenticity measures, and found that these scales had very similar and uniformly adaptive patterns of correlation with external variables, including agreeableness.We generated an item list with common instructions and response anchors to isolate content specifically reflective of realness as manifest in thoughts, feelings, motives, and behaviors.In the second study, we administered self and informant report versions of these items along with a number of criterion measures in order to select the best functioning indicators.We termed the resulting 12-item measure the Realness Scale (RS).In the third study, we confirmed the unidimensional structure of the RS and examined its correlations with other authenticity measures.In the fourth and fifth studies, we examined correlations between the RS and personality traits in large undergraduate and community samples, with particular attention to associations with agreeableness.In the sixth study, we used a peer nomination sampling strategy to dissociate realness from likability or positive valence.In the seventh study, we evaluated the retest and mean-level stability of the realness scale across five months.In the eighth study, we applied the RS as an observational measure to code speed-dating interactions, in order to evaluate the inter-rater reliability of observed ratings of realness.In the ninth study, we translated the RS to German and examined its validity for predicting actual behavior in a variety of hypothetical situations, as well as perceived impacts of realness for self and others.Statistical significance was set at p < .01 for all statistical tests, and interpretation largely focused on effect sizes.No study was preregistered.All studies were approved by local IRB boards and all data are available at https://osf.io/3kqpw/?view_only=2c91999e89c5457a9b5817fc8efc615d.

Study 1
The goal of the first study was to identify realness within the content of popular authenticity measures, and to establish associations between these measures with a variety of criterion variables.We administered three commonly used measures to 983 undergraduates (mean age = 19.46,SD = 2.07; 776 female/207 male; 713 white, 79 black, 138 Asian/Pacific Islander, 51 multiracial, 2 other; 41 Hispanic/941 non-Hispanic/1 unreported) and examined the intercorrelations and external correlates of their scales.

Measures
The Authenticity Inventory, Version 3 (AI; Goldman & Kernis, 2004) is a 45-item measure of authentic personality style with four dimensions (Kernis & Goldman, 2006): awareness (Cronbach's a = 0.77), unbiased processing (a = 0.66), behavior (a = 0.73), and close relationships (a = 0.76).The Real Relationship Inventory (RRI; Gelso et al., 2005;Kelly et al., 2010) is a 24-item questionnaire designed to measure two characteristics of a real relationship relevant for psychotherapy (but applicable to any relationship), realism (having realistic appraisals of self and other, a = 0.92) and genuineness (a = 0.91).The Authenticity in Relationships Scale (AIRS; Lopez & Rice, 2006) is a 37-item questionnaire designed to measure two relational authenticity factors (Gouvea et al., 2013): unacceptability of deception (a = 0.89) and intimate risk taking (a = 0.89).Both the RRI and AIRS instruct respondents to rate their behavior in a specific relationship.We asked people to respond to RRI and AIRS items in terms of their relationships with friends.
We administered a variety of criterion measures of adaptive and maladaptive personality.We generally expected authenticity scales to correlate positively with adaptive personality traits.The Five Factor Model Rating Form (FFMRF; Mullins-Sweatt et al., 2006) is a 30-item measure of the five-factor model of personality with one item for each facet of the NEO-PI-R (Costa & McCrae, 1992).We used the FFMRF to measure five personality domains: neuroticism (a = 0.68), extraversion (a = 0.71), openness to experience (a = 0.71), agreeableness (a = 0.69), and conscientiousness (a = 0.80).The Inventory of Interpersonal Problems-Short Circumplex (IIP-SC; Hopwood et al., 2008;Soldz et al., 1995) is a 32-item measure of interpersonal dysfunction.The circumplex octant scales (Mdn.a = 0.76) can be summarized with three factors: overall

Results and discussion
The top section of Table 2 shows that the intercorrelations among the various authenticity scales tended to be high, particularly for the two RRI scales (r = 0.92).However, there was some variation in these correlations, pointing to a distinction between more internal and external aspects of authentic behavior.For instance, the unbiased processing scale had relatively weak correlations with scales explicitly focused on external behavior such as AIRS intimate risk taking (r = 0.17).
The middle section of Table 2 shows that these scales also had a highly similar pattern of association with external variables, largely indicative of psychological health and well-being.Specifically, authenticity scales tended to have positive correlations with extraversion, openness, agreeableness, conscientiousness, and internal locus of control and negative correlations with neuroticism, interpersonal problems, sensitivity to others' warmth, borderline, dependent, and narcissistic personality, depression, alexithymia, and attachment avoidance and anxiety.We computed Pearson congruence coefficients1 to examine the similarity of these correlation patterns.These values, shown in the bottom section of Table 2, ranged from 0.77 to 1.00, with the majority > 0.90.Overall, these values suggest that, despite their different conceptual interpretations, each authenticity scale across these three multidimensional instruments has a very similar pattern of external correlates.Overall, these results indicate that authenticity is generally adaptive and each instrument is tapping a very similar construct, although there may be some subtle distinctions between more internal and more external aspects.
This pattern is somewhat consistent with what we would hope for from a measure of realness, but there were also notable divergences.The fact that all authenticity scales are highly intercorrelated suggests that the construct coheres at a broad level, as we would expect.
Validity correlations suggest that authenticity is a generally adaptive construct, and in fact it is strongly related with a number of indicators of mental health.However, the tendency to be agreeable and sensitive to others' coldness was among the strongest correlates of existing authenticity scales.As discussed above, these results do not align with our concept of realness, which is rooted in theories of authentic personality and behavior that have consistently prioritized genuineness over social grace, politeness, or impression management.Overall, these results were consistent with our hypothesis that the core of authenticity lies in the connection between its more internal and external features, but that this core is masked in existing measures by positive valence, efforts to assess internal and external dimensions as distinct from one another, and the influence of peripheral factors.

Study 2
Although these issues are difficult to disentangle at the level of existing subscales, we observed that some specific items seemed to reflect our concept of realness whereas others did not.The purpose of Study 2 was to isolate realness content within existing authenticity measures in order to generate a reliable, unidimensional measurement tool.We focused on two specific features of items within existing measures.First, items should reflect external behaviors that align with internal states.For instance, they should reference being transparent, open, and straightforward in social situations.In contrast, they should not solely reference internal or external states (e.g., self-awareness).We especially preferred items that specified the situations in which realness would be observed -that is, when there were potential costs to being real.This was based on our conceptualization that realness involves being authentic even when there are downsides.
Second, because we were interested in realness as a generalized individual difference that would be relatively stable across situations and relationships, items should refer to a general trait rather than a specific behavior or certain relationships.We had two reasons for this choice.First, although it is widely understood that authenticity is more likely and, arguably, appropriate in some kinds of relationships than others, we defined realness as a stable trait whose between-person rank order would generally pertain across situations with different presses for being real.Second, we sought to develop a general measure that could be used for a variety of purposes in the pursuit of understanding realness, rather than a measure that would be useful only for studying certain kinds of relationships.
Three of the authors independently identified items that best reflected these features on the AIRS, RRI, AI, and Wood et al. ( 2008) Authenticity Scale items.Of the 118 total items on these four instruments, we found 42 non-redundant items reflecting our notion of realness.However, these were not balanced with regard to thoughts, feelings, behaviors, and motives/desires, nor in terms of item keying.Thus, we wrote 38 additional items (all 80 items are available at https://osf.io/7vdgp/?view_only = 2 c91999e89c5457a9b5817fc8efc615d).We harmonized all items with a 4-point scale (False, Somewhat False, Somewhat True, True) and reworded items to reflect a general disposition rather than specific relationships as necessary.We then collected data with the goal of trimming these 80 items in order to construct a unidimensional realness scale that we could use to examine selfinformant agreement and correlates.
In addition to the 80 candidate realness items, participants also completed the following criterion measures.The Five-Factor Model Rating Form (Mullins-Sweatt et al., 2006) is a 30-item measure of the personality traits neuroticism (a = 0.75), extraversion (a = 0.70), openness (a = 0.63), agreeableness (a = 0.68), and conscientiousness (a = 0.80).They were also administered the honesty scale from the HEXACO-60 (Ashton & Lee, 2009, a = 0.73) and scales from the Computer Adaptive Test of Personality Disorders (Simms et al., 2011) measuring emotional detachment (7 items a = 0.85), manipulativeness (6 items a = 0.89), and submissiveness (6 items a = 0.89).The IIP-SC was used, as in study 1, to assess interpersonal problems and its octant scales (Mdn.a = 0.76) were summarized in terms of overall interpersonal distress, problems related to dominance vs. submissiveness, and problems related to warmth vs. coldness.Finally, the Self-Monitoring Scale was administered (a = 0.60).We expected realness to correlate negatively with self-monitoring.Informants completed the 80 realness items with reference to the target (participant), as well as the FFMRF Note.p-values and CIs are not given for ease of presentation.(neuroticism a = 0.75, extraversion a = 0.69, openness a = 0.57, agreeableness a = 0.73, conscientiousness a = 0.85).
Items were selected based on a consideration of several criteria (Holden & Fekken, 1990).First, we preferred items with a relatively low reading level.Second, we sought to balance reverse-and positively-keyed items, to minimize the potential impacts of response style (Nunnally & Bernstein, 1994).For negativelykeyed items, we focused on not being real to avoid some personal or social cost.Third, we selected items with strong part-whole correlations, to be sure we were targeting items at the core of the realness construct.Fourth, we sought balance in item content in terms of sharing thoughts, sharing feelings, sharing desires, and showing authenticity in behavior.

Results and discussion
Based on a consideration of each of these factors, we ultimately selected the 12 items in Table 3 for the Realness Scale (RS).These items are balanced in terms item keying and content (i.e., three items explicitly refer to thoughts and feelings each, four to motives or desires, and two to behaviors).Eight of the items explicitly reference a potential social tradeoff of being real.The Flesch-Kincaid reading level for all items was 7th grade or lower with a median of 4.5 and part-whole correlations (with reference to all 80 initial items) were all above 0.3.The alpha was 0.81 and the item mean was 2.78 (SD = 0.56).We fit these items to a factor model with diagonally weighted least squares (WLSMV) estimation to deal with categorical indicators using Mplus version 8 (Muthén & Muthén, 1998-2017).A model with one substantive factor and one method factor where the reverse keyed items were freed showed adequate fit to the data (X 2 (48) = 153.79,RMSEA = 0.05, CFI = 0.96, TLI = 0.95, SRMR = 0.03).Although one item (''others might see me as fake") had a relatively small contribution to the general factor, we retained it to enhance content validity.Specifically, this is one of the two items asking about other peoples' impressions of the respondent's behavior.
The informant version of the RS had an alpha of 0.83 (M = 3.03, SD = 0.53).The self-informant correlation for the RS was 0.16 (p = .002;95% CI = 0.06 -0.26), lower than we had expected.Table 4 displays correlations between realness, as rated by both self and informant, and big five traits, as rated by both self and informant.Self-reported realness correlated positively with self-reported extraversion and conscientiousness, negatively with neuroticism, and was uncorrelated with agreeableness and openness.Informant-informant correlations mirrored this result, but with somewhat stronger correlations with openness and agreeableness.The only significant correlation between informant-reported realness and self-reported traits was for high extraversion.No correlations between self-reported realness and informant-reported traits were significant, although there was a potentially meaningful negative correlation between self-reported realness and informantreported agreeableness.
Overall, these results suggest that, although agreement between self and informants is relatively low, similarity in the nomological networks of realness as assessed by self and informant was relatively high.The RS is generally associated with adaptive personality features.However, whereas by self-report people who see themselves as real do not also see themselves as agreeable and are not seen by others as agreeable, informants who rated targets as more real also perceived them to be more agreeable.This may suggest that realness is an attribute that is appreciated or experienced as agreeable or warm by others, even if it is not actually a particularly agreeable or polite pattern of behavior.We explore this issue further in Study 6 below.
We next examined correlations of self-and informant-rated realness with the measures that were only administered by self-report (Table 5).As expected, realness was correlated positively albeit modestly with honesty and interpersonal dominance and negatively with emotional detachment, manipulativeness, submissiveness, interpersonal problems, and self-monitoring.Notably, it was uncorrelated with warmth.Informant-rated realness had fewer significant correlations, likely due to method effects and lower power due to the smaller sample.Unlike self-reported realness and mirroring the effects for big five variables, it was positively correlated with warmth.It was also negatively correlated with emotional detachment and manipulativeness.
To summarize, in this study we were able to identify 12 realness items that formed a unidimensional scale, were balanced in keying, and covered a range of content involving being real with others about thoughts, feelings, desires, and behavior.Correlates between this scale and self-reported criteria were as predicted.However, self-other agreement was comparatively low, perhaps because participants and informants seemed to understand realness somewhat differently.Both self-and informant-reports associated realness with healthier emotional and interpersonal functioning.However, individuals considering their own behavior associated realness with assertive dominance, whereas individuals considering a target's behavior associated realness with warm agreeableness.

Study 3
The goal of Study 3 was to confirm the unidimensional structure of the RS and examine associations with existing authenticity measures, in order to confirm that the scale captured an essential feature of this complex construct.We administered the RS and five authenticity measures to 504 Amazon Mechanical Turk workers (mean age = 36, SD = 11.32;255 female, 245 male, 4 other; 353 white, 35 black, 7 Asian/Pacific Islander, 41 Hispanic, 68 other) who were paid $7/hr for participating in the Fall of 2016.Authenticity measures included the three that were used in Study 1 (AIRS, a = 0.95, RRI, a = 0.96 and AI, a = 0.94) as well as the Wood et al.
(2008) Authenticity Scale (a = 0.89) and three items adapted from Fleeson and Wilt (2010, Study 2; a = 0.92).The mean score on the RS items was 2.78 (SD = 0.56).Covariance among the items fit a model with a method factor well (X 2 (48) = 138.87,RMSEA = 0.06 [CI = 0.05-0.07],CFI = 0.96, TLI = 0.95, SRMR = 0.04).Correlations between the RS and established authenticity measures ranged from r = 0.47 to 0.64 (Table 6), suggesting that the RS is strongly associated with existing authenticity measures, even though its content is more specific to realness as we define it.
Results are given in Table 7.As in Study 4, realness was negatively correlated with neuroticism and positively correlated with extraversion, openness, and conscientiousness.Contrary to our hypothesis, a modest but significant positive correlation was observed with agreeableness.Overall, the results of Studies 3, 4, and 5 largely confirm our expectation that the RS is a) positively and strongly correlated with existing authenticity measures, b) like other authenticity measures in having generally positive correlates, but c) different from other authenticity measures in being mostly unrelated to agreeableness.However, given the importance of the agreeableness effect in particular, and the ambiguity in these results (i.e., a significant agreeableness correlation in Study 5), we next sought to pursue the association between realness and agreeableness in more detail.Specifically, we designed a study to disentangle perceptions of a people who are equally likeable but differ in realness.

Study 6
In Study 6, we used a peer nomination strategy to test whether informant report correlations with agreeableness traits were due to valence or selection effects, and to dissociate realness from likability or positive valence more generally.Specifically, we asked people to select two friends, both of whom were equally likeable, but one of whom was particularly real whereas the other was not.We reasoned that this design would eliminate the positive association between realness and agreeableness from informantreports (Table 4), and thus show that realness can come across as relatively disagreeable when referenced against a more courteous (but equally likeable) alternative.
Participants were 746 undergraduates in a public American university who were compensated with course credit.The average age was 19.81 (SD = 2.10); 625 were women, 117 men, and 4 reported other genders; 372 were Asian/Asian-American, 165 White/ European-American, 12 were Black/African-American, 11 Pacific Islander, 5 Native American, 68 multi-racial and 113 reported other races or did not report race; 179 participants identified as having Hispanic ethnicity.Participants were given the following instructions (order of second and third paragraph was randomized): ''Please think of two friends that can be described by each of the following paragraphs.You should like these two friends about the same, and feel about equally close to both of them.
The first friend is not that concerned about seeming rude, and others see them as 'real', direct, straightforward, and even brutally honest.They say what they want rather than what others want to hear and tell others what they think even if it causes conflict.They do what they feel or want whether or not others want or feel the same thing, even it makes the situation awkward or uncomfortable.They tell the truth even if it makes others unhappy and are not afraid to tell someone else if they feel hurt or disappointed.This person would never act like they believed something others believed if they did not feel it themselves.
The second friend is very concerned about seeming rude, and others see them as polite, discreet, flexible, and possibly even 'inauthentic'.They say what they think others want to hear rather than what they actually want in order to avoid conflict.If they believe others want or feel something strongly, they may pretend that they want or feel the same thing to keep the situation comfortable.They may not tell the whole truth if they fear it will make others unhappy, and would avoid telling someone else they feel hurt or disappointed.This person would be willing to act like they believed something others believe to avoid an awkward situation." They were then asked to rate the personality characteristics of each of these friends (order of friends was randomized) using the IPIP-120 (Maples et al., 2014, median facet a = 0.75 for real friend and a = 0.75 for polite friend), the Agentic and Communal Values scale (Trapnell & Paulhus, 2012; alphas ranged from 0.78 to 0.94), and the Honesty scale from the brief HEXACO-60 (Ashton & Lee, 2009; a = 0.78 for real friend and a = 0.81 for inauthentic friend).
The profile correlation for mean personality facet scores across the ''real" and ''inauthentic" friend was r = 0.22, suggesting rather different profiles for these two vignettes.Participants rated the real friend as substantially more extraverted, assertive, and agentic and substantially less agreeable, communal, and cooperative (Table 8).As with self-report correlations, peers rated their friend who is more real as less neurotic, but in contrast to other findings, they also rated that person as less conscientious.Interestingly, raters actually perceived the inauthentic person as more ''honest" based on the HEXACO scale, highlighting both differences between a self-report and peer-nomination approach to studying authenticity-related constructs, and substantive differences between realness and HEXACO honesty.
Overall, these results suggest that the informant-report correlation between realness and agreeableness observed in Study 2 may have been, at least in part, an artifact of the general halo effect associated with informant-reported personality traits (Leising et al., in press).That is, when asked to describe a friend, there is a tendency for people who say positive things in one domain (i.e., realness) to extend those positive descriptions to other domains (i.e., agreeableness).However, when asked to distinguish real and polite friends that are equally likeable, people tend to see the real friend as relatively less agreeable.Self-report data tend to be between these two effects in that the association between realness and agreeableness is small or null.This finding supports our contention that a critical element of being real involves the ability to be disagreeable at times, and highlights the impact of different approaches to asking people about this kind of behavior.

Study 7
The goal of Study 7 was to estimate the retest stability of realness.We sampled 412 undergraduates (mean age = 20.35,SD = 2.13; 81% women, 17% men, 2% other; 36% white, 34% Asian/Asian-American, 3% black, 2% Native American/Pacific Islander, 25% other).Of these, 301 completed the RS five months after baseline.The retest correlation was 0.74, and the mean among people who participated at both waves was 2.45 (SD = 0.53) at time 1 and 2.45 (SD = 0.52) at time 2 (d = 0.00).Overall, these data support the interpretation of realness as a stable individual difference variable, with longitudinal consistency estimates similar to what we would expect of personality traits.

Study 8
The goal of Study 8 was to test whether realness could be reliably coded in actual behavior by observers.Part of the logic of this study is that traits can be inferred from the observation of multiple contextualized states (Lenton et al., 2013), and that observations of individuals across multiple interactions produces more valid estimates of realness than those from a single social situation (Gershon & Smith, 2020).Thus, we had trained research assistants observe targets across 12 interactions with different people, and then use the RS scale items to estimate how real the person was in their interactions, in general.
Specifically, eight research assistants (four women and four men) coded interactions during 8 speed dating sessions with 24 people each, for a total of 192 targets (Eastwick et al., 2007;Finkel & Eastwick, 2008).Each session was coded by 4 coders (2 men and 2 women), who watched all 12 interactions for a given participant and then filled out the 12-item RS based on what they think the person is like, in general.The overall alpha for the RS was 0.90, and the one-way random effects average ICC for the overall realness score was 0.59 (95% CI = 0.48 -0.68; individual ICC = 0.26, 95% CI = 0.19 -0.35), indicating statistically significant (p < .001)agreement across observers about individual differences in realness.We explored correlations between observer-rated realness and a brief self-report measure of big 5 traits (Donnellan et al., 2006; neuroticism a = 0.78, extraversion a = 0.84, openness a = 0.65, agreeableness a = 0.75, conscientiousness a = 0.74).All correlations were < |0.05| and none were statistically significant, perhaps because of method effects.

Study 9
The goals of Study 9 were to translate the RS to German, evaluate correlations with measures of personality and interpersonal effectiveness, and test the ability of the RS to predict behavior in hypothetical interpersonal scenarios in which individuals would have to decide between either being real or not in situations in which either response may come with risks.As a first step, the items from the RS were translated by two bilingual researchers, who then agreed on consensus items.The items were then back translated by a third bilingual speaker, and checked for accuracy against the original items.The German version of the measure was then given to a sample of 204 participants from the general population (142 women, 62 men) with an average age of M = 31.3years (range = 18 to 71, SD = 13.13).A model with one substantive factor and one method factor where the reverse keyed items were freed showed adequate fit (X 2  9.These results essentially replicate studies reported above in English speaking samples that showed positive correlations between realness and extraversion and negative correlations with neuroticism.Correlations with openness and conscientiousness were positive but not significant at p < .01,and the correlation with agreeableness was negative but weakest in absolute value among the big five traits.The results add to our previous findings by showing that RS is related positively to interpersonal competence and negatively related to fear of negative evaluation.Given the similarity of empathy to agreeableness, we were not surprised to find realness was not significantly correlated with empathy. Study 9 participants were also asked to respond to nine scenarios in which they could be real or not, in which either choice could have a perceived cost.For each scenario there were two different versions, one highlighting costs for self and the other highlighting costs for others.The participants were divided randomly into two subsamples (N = 107 and 97), for which the scenarios either emphasized cost for the self or for the other.For instance, in one scenario, the participant is told that they have accidentally shared a friend's secret, and are asked whether they tell the friend about this.In one version of this scenario, the emphasis is placed on the respondent's discomfort discussing this with their friend (cost for self), whereas in the other, emphasis is placed on how the friend may feel hurt (cost for others).The respondents were then asked how likely they would be ''real" on a 4-point scale (in this case, tell the friend).Alphas were 0.60 (self) and 0.61 (other) for these scenarios.The RS correlated r = 0.58 and r = 0.55 for the self and other scenarios, respectively, strongly supporting the validity of the RS to predict reports of contextualized social behavior.

Discussion
At the moment, the world is awash in ''fake news", citizens are routinely manipulated by politicians who do not mean what they say, and social media platforms incentivize virtue signaling and punish straightforwardness.Although being ''yourself" is often extolled in modern society, it comes with social risks.It is these moments of social risk that provide perhaps the most valid test of whether a person is actually being real: a person who is only real when it pays off is not really real at all.This complexity is emphasized in classical psychological theories about authenticity and related concepts (congruence, genuineness, transparency), yet contemporary research uses measures that are strongly related to agreeableness, and which tend to mix content that is central to authentic behavior with content that is more peripheral.We sought to identify, distinguish, and validate the tendency to be real, the core individual difference variable underlying authentic personality processes, which we define as doing on the outside what one feels on the inside regardless of the proximal social consequences.
Realness may be a particularly important individual difference variable within certain domains of social behavior.For instance, being real may be both harmful and beneficial for politicians, but for citizens, is a key characteristic of trust (Rosenblum et al., 2019).As such, both actual demonstrations and (potentially inaccurate) perceptions of realness are nearly always an important consideration in the political sphere.Related, standing up to or criticizing powerful people and institutions to promote social justice is socially risky, by definition.People who have been made famous for doing so (e.g., Joan of Arc, Sitting Bull, Colin Kaepernick, Thomas Paine, Rosa Parks, William Tell, Henry David Thoreau) strike us as prototypically real -and they have historically experienced both the costs and benefits of this trait.To the degree that being real is an important ingredient for making the world a better place, understanding and promoting realness at the individual level may contribute to a more just society.At the same time, people who both hold and express hateful, racist, and divisive beliefs are also being real.As such, the social value of realness may depend on the health of those inner qualities that support it, such as selfawareness and capacity for reflection.
Realness may be particularly important in close relationships, such as psychotherapy, romance, or parenting.Indeed, we would hypothesize that, all things equal, most people would rather have a close relationship with someone who is real than with someone who is not.Again, however, we would expect that realness would be particularly valued in close relationships when it is supported by internal capacities for empathy and personal reflection.This notion is captured by the idea that people generally prefer a friend whose ''heart is in the right place".
These speculations point the way to future research that will benefit from our generation of a unidimensional model of realness.In these studies, realness was relatively stable, observable, predictive of contextualized social behavior, positively associated with adaptive functioning, and largely unrelated to concerns about being agreeable vs. antagonistic, as predicted.These results have implications for understanding individual differences in an important pattern of social behavior and may help clarify disconnections between classical theories and contemporary research on authenticity.

Realness and authenticity
Authenticity has captured the attention of theorists and researchers for decades, but it is a highly complex construct that has proven difficult to study and around which no scholarly consensus has emerged (Hicks et al., 2019).The authenticity literature is somewhat disjointed, with measures that are similar but not identical, and in which theory and research have parted ways in important respects (Baumeister, 2019).Moreover, our results suggest that existing measures deviate from classical theories about authenticity in being strongly related to agreeable personality characteristics.
Based on our literature review, we concluded that this was a result of two main factors.The first was that existing measures seem to capture some non-specific social desirability variance that contributes to discriminant validity issues with respect to agreeableness-related traits and behaviors.The second was the effort to account for multiple internal and external features that give rise to authentic behavior, even if they are supportive but not essential.We understand authenticity as a relatively complex, multi-component, within-person process involving dynamic connections between internal states and external behavior.Many of the existing authenticity measures were based on theories that explicitly referenced such dynamic, multi-component, within-person processes.These processes included some features that seem central to authenticity (behavioral expressions of inner states), as well as other features that may support authentic behavior but in a somewhat non-specific way (e.g., self-awareness).
To be clear, we think that studying authenticity and all of the processes that support it is an important endeavor for social scientists.However, we concluded that, rather than trying to capture all of the features involved in complex within-person authenticity dynamics using measures designed to detect between-person differences, it would be better to begin by isolating a core betweenperson variable that is central to authentic behavior.A firm model of individual differences in realness can help facilitate authenticity research by distinguishing those individuals most likely to be real in a given situation, and by providing a variable that can be used to study the within-person contours of real behavior across time and situations.
We found that realness content was present in existing multidimensional measures of authenticity, but that it was also obscured in measures with scales that focused on either internal characteristics such as capacities for personal awareness, accurate perception, and reflective function, or external characteristics involving explicit social behavior.While such characteristics, in combination, may support authenticity, it is not being aware or behaving in a certain way in isolation that provides evidence that someone is authentic -it is the correspondence between these inner and outer states.This correspondence could be labeled congruence or transparency, terms which directly indicate the connection between inner and outer states.However, the second obscuring factor was that item content on existing measures tended to have a strong positive valence.A consequence of this positive valence is that authenticity measures tend to be strongly correlated with agreeable traits.However, as described in detail above, this pattern of correlation departs significantly from classical theories of authenticity.An authentic person should be so whether or not there are potential negative consequences.In fact, situations in which the potential for negative consequences are present provide the truest tests of authenticity.We refer to this tendency to be transparent or congruent without regard for social consequences as realness.By realness, we simply mean that when a person reveals everything they think, feel, and want on the inside to others in a way that is direct and straightforward, they are being real; when they conceal such features, they are being fake.
To be clear, realness does not solve all of the problems with authenticity.A significant hurdle is that the validity of realness scores depends on the rater having a valid account of inner states.Generally speaking, the self is the best source of information about inner states, although individuals may have not accurately report them for a variety of reasons.Observers and informants, in contrast, may not share all of the self's blind spots, but they also do not have direct access to the target's inner states.It may be possible to create experimental approaches to test the relevance of self-insight to some degree (e.g., by manipulating inner states directly via priming techniques), which would be an important direction for future work.
One specific way in which realness may be different from authenticity occurs when a person has two motives.For instance, a person may disapprove of someone else's behavior but also value social harmony, and expect that expressing that disapproval would create disharmony.It is not clear whether expressing disapproval or not would be the most authentic behavior in this situation.However, the most real response would be to both express disapproval and also express the desire to maintain social harmony.To the extent that either of these inner states or motives are concealed, the response not real (but still could potentially be authentic in at least some sense).Future work focused on the how people express themselves when their motives conflict would be informative about both realness and the broader concept of authenticity.

Correlates of individual differences in realness
We found that individual differences in realness were strongly related to variation in existing measures of authenticity and correlated with high levels of extraversion, openness, conscientiousness, honesty, dominance, internal locus of control, and interpersonal competence.Realness was negatively associated with neuroticism, a range of maladaptive personality characteristics, interpersonal problems, self-monitoring, and fear of negative evaluation; and it was largely unrelated to agreeableness, although the pattern of results was complicated, as we will discuss in more detail below.Overall, this pattern of correlations suggests that people who are more real tend to have more adaptive personalities.This is consistent with classic theories that postulate that realness is an outgrowth of psychological maturity (e.g., Horney, 1951;Maslow, 1968).However, as discussed above, this may depend on the level of health of inner characteristics such as self-awareness and capacity for reflection and emotion regulation.In other words, it may be the case that realness is adaptive among healthy, prosocially motivated individuals, whereas it is maladaptive or even pernicious among people who are less well-developed or antisocial.Indeed, we note that children are often seen as characteristically ''real", despite not having developed personalities.Given that both classical theory and our data imply but do not prove that realness is an outcome of healthy maturation, genetically-informed developmental data would be useful for better understanding the sources of individual differences in the construct (Wagner et al., 2020), and future research should seek to distinguish being real from the healthy inner capacities that support personal and interpersonal adaptation.
Although we conceptualize realness as an individual difference construct, we also wish to emphasize that it is importantly different from the big five or analogous personality traits.Personality traits such as those in the big five indicate the tendency to behave in a certain way, relative to others, across time and situations.For instance, people who are high in extraversion are more extraverted than most other people in most situations.In contrast, realness is a contingent construct, in that it is only possible to test whether someone is real when social risk is present.As such, it is most telling to observe realness when the relevant costs are present.In an individual difference measure such as the RS, this can be specified in the items themselves.In observational or experimental work, this would have to be taken into account in other ways, such as the manipulation of scenarios so as to create social risk.This would be a fruitful avenue for future research because it would help inform the mechanics of real behavior, and help distinguish it from other kinds of traits.

Realness and conceptually similar constructs
Some of the modest correlations between realness and conceptually similar constructs are important for understanding the difference between realness and other aspects of authenticity.For example, honesty as conceptualized on the HEXACO is a relatively instrumental trait with significant positive valence (e.g., If I knew that I could never get caught, I would be willing to steal a million dollars (reverse), I wouldn't use flattery to get a raise or a promotion at work, even if I thought it would succeed).In contrast, the social costs of realness are embedded in the items of the RS, which also focus on being real for its own sake, as opposed to the instrumental utility of the alternative.To be concrete, HEXACO honesty might be better at capturing the tendency (not) to use subterfuge in order to get something or impress someone, RS realness might be better at capturing the tendency to act according to inner experience regardless of personal or social consequences.
Self-monitoring is another conceptually similar but somewhat broader and empirically distinct construct.Self-monitoring focuses on behavioral expression, and particularly non-verbal expressions (Snyder, 1974).Moreover, the absence of self-monitoring can function to be either real or non-real.For instance, according to Snyder (1974), one of ''the goals of self-monitoring may be to communicate accurately one's true emotional state".In other words, for a person who is characteristically deceptive or fake, an absence of self-monitoring would tend to contribute to being less real.Overall, we see self-monitoring as capturing some aspects of being real in the sense that the absence of self-monitoring is thought to produce a tight, non-reflected connection between internal states and outward behavior, but that the concept also includes some of the internal features depicted in Fig. 1, and may not necessarily be associated with being real in any particular situation.The relatively modest correlation between realness and self-monitoring in study 3 is consistent with this interpretation.
Disinhibition, a third conceptually similar construct, is a broad trait involving impulsive behavior.It tends to be associated with negative outcomes such as externalizing psychopathology (Patrick et al., 2013), and tends to decrease normatively with age (Vaidya, Latzman, Markon, & Watson, 2010).There is a similarity between being real and being disinhibited, because both of these concepts involve a connection between inner states and behavioral expression.However, disinhibition is broader and more maladaptive, and thought to reflect a kind of psychological immaturity or underdevelopment.For instance, whereas disinhibition is a strong predictor of substance use (Iacono, Malone, & McGue, 2008), we would not expect realness to be related to substance use.Instead, we would expect people who are real to use substances if they feel like using them, and not use substances if they don't, whereas we would expect disinhibited people to experience an urge to use substances that they find difficult to control.Disinhibition has been conceptualized as low conscientiousness (Clark & Watson, 2008); in this study the RS was consistently albeit modestly negatively correlated with conscientiousness, supporting the empirical distinction between realness and disinhibition.

Realness and agreeableness
One of the main motivations for this research was our observation that classical theories of authenticity emphasized the potentially disagreeable aspects of realness (e.g., Maslow, 1968) whereas existing measures of authenticity had uniformly positive correlations with individual differences in agreeable behavior (e.g., Pinto et al., 2012).We concluded that this discrepancy may be due, at least in part, to social desirability.Generally speaking, authenticity and agreeableness are both positive characteristics, and thus items designed to assess them might contain non-specific positive valence, creating a correlation between the two constructs (Baumeister, 2019;Jongman-Sereno & Leary, 2016).
Comparisons of validity correlations from self, informant, and peer-nomination data were used to disentangle social desirability effects.The self-report correlation between realness and agreeableness was negligible.The correlation between informant-rated realness and informant-rated agreeableness was positive, which may suggest that informants would generally prefer their friends to be real.This interpretation is consistent with assertions by theorists like Rogers (1961) regarding the interpersonal importance of being real.However, when given a forced choice between a real and a polite friend, both of whom the rater likes, informants rated the polite friend as substantially more agreeable than the real friend.This pattern can be summarized as follows: people who are more real do not tend to see themselves as more agreeable, but people tend to see realness in their friends as more agreeable than otherwise, while also recognizing that it is less agreeable to be real than to be polite.
Longitudinal and experimental work would be useful for further disentangling realness from disagreeableness, from the perspective of both the self and others.Further refinement of the measurement of these constructs may also be useful.Specifically, it may be that realness is experienced as warm or communal in a deep sense, even if it is not agreeable in the more superficial sense.Colloquially, people often experience gratitude when others are ''real" with them, presumably because they attribute that realness to some kind of deep or lasting concern.Given the possibility that perceived agreeableness and realness reflect different levels of psychological functioning, it may not make sense to measure them with the same kinds of tools (Leary, 1957), and it may be profitable to develop techniques that distinguish deeper, motivational aspects of behavior from more visible, superficial aspects.

Realness, context, and states
One interesting finding from recent research is that people tend to report feeling more authentic when they are their best selves, not their most typical selves, in social situations (Beer & Harris, 2019;Fleeson & Wilt, 2010).This speaks to the valence effect discussed above -people want to believe they are their best selves deep inside, which includes being authentic (Hicks, Schlegel, & Newman, 2019), and there is a fairly consensual model of what the best self is (Bleidorn et al., 2019).This may help explain why ratings of authenticity and ratings of adaptive personality traits, including agreeableness, converge at a very general level.
But a different and perhaps more interesting behavioral question is, in the moment when the crisis strikes, are you real (Sedikides et al., 2019)?Being real in this sense is not the same thing as behaving according to one's typical trait levels, being the same way across all situations, or being the best version of yourself.As inner feelings may change dramatically across situations or roles, then behavior must correspondingly change, given that realness is defined by the congruence between inner and outer states.Realness is consistency with how one feels in a given moment, which itself might change across situations, and which may deviate from typical traits.A related question is, what if a person has an internal conflict and their behavior only corresponds to one side of that conflict?We would argue that this would be only partly real, and to be fully real, one should outwardly express both aspects of their internal conflict.
Longitudinal and contextualized, multi-method data are needed to test these kinds of hypotheses.We did not consider how contextual factors such as relationship closeness or hierarchy (Chen, 2019), the match between internal and external states (Eastwick, Finkel, & Simpson, 2019), relationship dynamics (Finkel, 2019), internal conflict (Strack & Deutsch, 2004), or the level of support in the environment (Ryan & Ryan, 2019) affect realness.We anticipate that, like other traits, realness will be strongly impacted by both individual differences and situational dynamics.In this set of studies, we focused on individual differences and learned very little about situational dynamics.By generating a valid measure of realness that can be administered as a self-report, informantreport or behavioral observation tool, we have we have provided a method for capturing this core feature of authentic behavior and set the stage for work on the manifestation and dynamics of realness states in actual social contexts.

Limits to generalizability
These studies were conducted exclusively in WEIRD samples in two countries.It would be important to examine how well the concept of realness generalizes to other cultures in terms of content validity, measurement invariance, and patterns of correlation before generalizing these results to people, in general.Even within these countries, efforts were not specifically made to examine how realness functions across important sub-segments of the population (e.g., different ethnicities or social classes).This is a related and important area for future work.It seems plausible that, within WEIRD countries, people with different backgrounds are more likely to exhibit realness than others.For instance, it may be that people with more historical or personal privilege experience relatively less social risk in being real than people from underrepresented or underprivileged groups.Extending from this idea is the possibility that certain known groups might be particularly high (e.g., counselors) or low (e.g., thieves) in realness.Studies sampling such groups would provide a novel means of validating and studying realness.

Conclusion
Our goal was to establish the construct of realness, or the tendency to act on the outside the way one feels on the inside regardless of social consequences.This is in contrast to fakeness, in which a person conceals certain aspects of the inner experience.A person who hides their thoughts, feelings, or motives from others is plainly not being real.Realness has mostly been studied in the context of authenticity, a complex within-person process often measured with tools that do not align in certain ways with classical theories.In particular, these measures try to capture a multicomponent temporal process using cross-sectional questionnaires, and to be saturated with positive valence, including agreeableness.As such, realness is somewhat obscured in contemporary authenticity research.In this study, we isolated realness within the general domain of authenticity, developed a tool that could be used to measure relatively stable individual differences in realness from the perspective of the self, informants, or observers, and established its properties.This sets the stage for future work on authentic social processes, and in particular the sources, correlates, costs, and benefits of being real.

Table 1
Hopwood et al., 2011003)roblemsrelated to dominance vs. submissiveness, and problems related to warmth vs. coldness(Gurtman & Pincus, 2003).The Interpersonal Sensitivities Circumplex (ISC;Hopwood et al., 2011) is a 64-item measure of behaviors that the respondent finds irritating in others.Like the IIP-SC, the circumplex octant scales (Mdn.a = 0.80) of the ISC can be summarized with three factors: overall interpersonal sensitivity, being annoyed by dominance vs. submissiveness, and being annoyed by warmth vs. coldness.The Hyperbolic Temperament Questionnaire (HTQ; Fraley et al., 2011ylor, 1994Scale of Locus of Control (LOC;Levenson, 1981; a = 0.59) is an say 8-item measure of internal vs.external locus of control.The Toronto Alexithymia Scale (TAS;Bagby, Parker, & Taylor, 1994; a = 0.85) is a 20-item measure of alexithymia, or difficulties with emotional awareness.The Experiences in Close Relationships Scale (ECR;Fraley et al., 2011) is a 36-item measure of anxious (a = 0.94) and avoidant (a = 0.95) attachment styles.

Table 3
Realness Scale item properties in Study 2.

Table 4
Self and informant correlations between realness and big five personality traits in Study 2.

Table 8
Differences in peer ratings of a real and inauthentic friend.

Table 9
Correlations between RS and individual difference measures in a German sample.
:. Realness scale Please read the following statements and use the scale provided to indicate how accurate each statement is of you, in general.Scale: False, Slightly True, Mainly True, Very True At times I say what people want to hear rather than what I want.(r) I do what works best for the situation even if it is not how I feel.(r) Others might see me as fake.(r) It would take a lot for me to tell someone they have hurt me.(r) I tell the truth even if it makes others unhappy.I say what I believe even if people don't like it.I sometimes act like I believe what others believe.(r) I tend to tell others exactly what I think even if it causes conflict.I avoid sharing desires that others may not approve of.(r) I express my needs and desires directly.I share my feelings with others even if it upsets them.I tell people what I want even if they may not want the same thing.