Abstract
Notwithstanding a recent flurry of organizational research on the construct of “situational strength,” research on the other side of the coin—“personality strength”—has rarely been conducted in organizational settings, has been scattered across multiple disciplines, has been called different things by different researchers, and has not yet been used to test theoretical propositions paralleling those in recent organizational research on situational strength. In the present review, drawing from several disparate research literatures (e.g., situational strength, personality states, traitedness, cross-situational consistency, scalability, appropriateness, self-monitoring, interpersonal dependency, hardiness, attitude strength, and self-concept clarity), we (a) define personality strength and contrast it with personality trait, personality strengths (plural), and layperson conceptualizations of the terms “strong personality” and “weak personality,” (b) briefly discuss the history of research related to personality strength, (c) identify a common prediction, emanating largely independently from several literatures, regarding the interactive effect of personality traits and personality strength on behavior, (d) articulate three novel predictions regarding the impact of personality strength on within-person situational and behavioral variability, (e) develop three broad categories of personality strength operationalizations (i.e., statistical, content-general, and content-independent) and discuss potential interrelationships among them, (f) suggest “best practices” for operationalization, thereby providing an agenda for future research, and, finally, (g) discuss the practical implications of this work for human resource management.
Keywords
I could be well moved, if I were as you . . . But I am constant as the northern star [W]hat has yet to be studied is the other side of the strong versus weak situation coin—namely, the possibility of “strong” versus “weak” personalities.
Why do people behave differently in different situations? To be sure, the content and “strength” of the situations themselves play important roles in influencing behavior (Funder, 2006; Johns, 2006; Meyer, Dalal, & Hermida, 2010). Employees, for instance, are more likely to exhibit conscientious behavior in the workplace when the cost of errors is high than when it is low (Meyer, Dalal, & Bonaccio, 2009). The impact of the work situation on job performance is certainly not lost on organizational researchers, who study topics such as workplace stressors, job design, leadership, and organizational structure and climate. Moreover, some research results suggest that behavior is influenced more heavily by the situation than by personality traits (see, e.g., Fleeson & Gallagher’s, 2009, meta-analysis). Yet equating personality solely with personality traits may understate the role of personality in determining employee performance. People seemingly differ not only in their characteristic levels of behavior across situations but also in their characteristic levels of variability in behavior across situations. This latter aspect of personality is our focus in the current article. We aim to increase researchers’ understanding of the phenomenon and its effects.
To this end, we review a large number of extant but thus far disparate research literatures from a new “lens”: that of personality strength. This perspective—which we create as a parallel to situational strength (Meyer et al., 2010)—permits us to draw connections across literatures, develop novel propositions, and discuss implications for future research as well as practice in organizational settings. Specifically, in the present review, we (a) define personality strength and contrast it with personality trait, personality strengths (plural), and the layperson conceptualization of personality strength, (b) briefly discuss the history of research related to personality strength, (c) identify a common prediction that has emanated (largely independently) from several literatures, (d) provide an overview of three novel predictions engendered by a personality strength perspective that emphasizes parallels with situational strength (thereby contributing to extant research on person-situation interactions by personality psychologists and organizational researchers; e.g., Buss, 1977; Endler & Magnusson, 1976; Meyer et al., 2009; Tett & Burnett, 2003), (e) develop three broad categories of personality strength operationalizations (i.e., statistical, content-general, and content-independent), (f) discuss potential interrelationships among these three categories of operationalization, (g) suggest “best practices” for future operationalization and measurement, and, finally, (h) discuss the importance of a personality strength perspective for the practice of human resource management in organizations.
Construct Definition
Like Locke and Latham (2004: 395), we envisage strong versus weak personality as “the other side of the strong versus weak situation coin.” A strong situation is posited to yield similar behavior from all persons who find themselves in that situation. In other words, strong situations theoretically reduce variability in behavior across persons, thereby inoculating behavior from the impact of personality (Meyer et al., 2010). For instance, although unconscientious employees may normally be more likely than conscientious employees to work carelessly, neither unconscientious nor conscientious employees are likely to work carelessly when the cost of errors is high. Consequently, situational strength has been formally defined as “implicit or explicit cues provided by external entities regarding the desirability of potential behaviors” (Meyer et al., 2010: 122).
Strength, therefore, is an important characteristic of the situation or context (Meyer & Dalal, 2009). However, we contend that strength is also an important characteristic of personality. In parallel fashion to a strong situation, a strong personality is posited to encourage similar behavior from a person, regardless of the situation. In other words, a strong personality theoretically reduces variability in behavior across situations within persons, thereby inoculating one’s behavior from the impact of the situation. For instance, conscientious employees are likely to work carefully on average across situations; however, compared to conscientious employees with a weak personality, conscientious employees with a strong personality are likely to exhibit less variability around that high average carefulness score. Consequently, personality strength is formally defined here, in a manner similar to situational strength, as “the forcefulness of implicit or explicit internal cues regarding the desirability of potential behaviors.”
To further clarify the definition of personality strength, it may help to define what it is not. To that end, we contrast the current conceptualization of personality strength with (a) personality trait, (b) personality strengths (plural), and (c) the layperson conceptualization of personality strength.
Personality Trait Versus Personality Strength
The first key distinction is that a person’s personality strength is not isomorphic with his or her standing on a particular personality trait. We conceptualize the personality trait score as reflecting the direction and extremity (relative to a midpoint) of internal behavior-relevant cues associated with a particular domain of personality (e.g., agreeableness, conscientiousness). In contrast, we conceptualize the personality strength score as reflecting the consistency of those cues across situations. Consider, for instance, a set of agreeableness-relevant situations (Ten Berge & De Raad, 2002) in the workplace: situations in which employees are proven wrong by other employees, situations in which employees want a coworker to do something, situations in which employees are sitting in the break room drinking a cappuccino, and so forth. Employees’ agreeableness trait scores describe their average level of agreeableness across these situations, whereas employees’ agreeableness strength scores describe the consistency (i.e., lack of variability) with which they exhibit agreeableness across these situations.
The distinction between personality trait and personality strength is an important one, and we return to it subsequently in our discussion of historical thinking about personality strength as well as in our discussion of best practices for the operationalization of personality strength.
Personality Strengths (Plural) Versus Personality Strength (Singular)
In reaction to the purported overemphasis of psychology on human weaknesses, researchers in the areas of “positive psychology” and “positive organizational behavior” have begun to focus explicitly on human strengths: namely, those aspects of the human experience that are beneficial to the self, work organization, and/or society (Luthans & Youssef, 2007; Youssef & Luthans, 2007). In the case of personality, the positive psychology and organizational behavior movements emphasize positive—in the sense of having positive effects—personality traits, such as conscientiousness and core self-evaluations (King & Trent, 2013; Luthans & Youssef, 2007). Such personality traits are often (see, e.g., King & Trent, 2013) referred to as personality strengths (plural). Yet, similar nomenclature notwithstanding, the emphasis in the positive psychology literature on specific personality traits and the positivity of their effects is very different from our emphasis on personality strength (singular) as an explanation of within-person variability in trait-relevant behavioral predilections across situations. Moreover, as discussed in greater detail in the practical implications section of this article, personality strength is not an unambiguously “good thing.” When personality determines behavior irrespective of the situation, the effects can be harmful as well as beneficial.
Current (Formal) Versus Layperson Conceptualization of Personality Strength
To further explicate the current conceptualization of personality strength, it is useful to contrast the two poles of this construct—that is, the strong personality and the weak personality—with layperson conceptualizations. In layperson terms, a person with a strong personality dominates other people and influences their behavior (e.g., Scheufele & Shah, 2000). Influencing the behavior of others could certainly be a way of proactively shaping situations to be more homogenous in terms of behavioral requirements. This, however, is a potential effect, not a necessary component, of strong personality as we define it. By maintaining that a strong personality is one who is extremely extraverted or dominant, the layperson conceptualization conflates personality traits with personality strength. On the contrary, an introverted employee could exhibit a strong personality by consistently behaving in a manner reflecting introversion across situations: for instance, by avoiding projects involving large amounts of client contact, by not talking a lot in team meetings, and by avoiding drawing attention to himself or herself.
Layperson conceptualizations of the construct’s other pole, weak personality, also differ somewhat from the current conceptualization. In layperson terms, a person with a weak personality is one who is dominated by other people, such that other people regularly influence his or her behavior. This would suggest that such a person exhibits great variability in behavior across situations as a function of the personalities of other people present (which constitute aspects of “the situation” encountered by the focal person). Behavioral variability across situations is certainly a hypothesized effect of our conceptualization of weak personality. However, the influence of other people is but one of many sources of situational strength (see, e.g., Meyer et al., 2010’s, review of the situational strength literature) and is therefore not a necessary cause of the behavior of people who we characterize as having a weak personality.
Foundations of Personality Strength
Most early thinkers did not distinguish between the concepts of personality trait and personality strength because they viewed behavioral consistency across situations as a sine qua non for the existence of a personality trait (Allport, 1937)—or even for “becoming a person” (Rogers, 1961). Early research did provide some dissenting findings: For instance, Hartshorne, May, Maller, and Shuttleworth (1928) found lower consistency than expected. Yet the issue remained largely dormant until Mischel’s (1968) provocative assertion that trait-based personality approaches had not demonstrated adequate cross-situational behavioral consistency. Specifically, Mischel suggested that the observed upper limit of personality–behavior correlations of around .30 reflected not just measurement error but rather evidence that behavior is influenced by situations to a greater extent than trait theorists were willing to concede.
The resulting defenses of trait-based conceptualizations by personality psychologists included several related ideas: individual differences in the consistency of responses to personality scales (Bem & Allen, 1974), individual differences in the extent of behavior changes related to equivalent situational changes (Lanning, 1988), individual differences in the relevance of various personality traits (Paunonen, 1988), and individual differences in whether a personality trait is possessed at all (Baumeister & Tice, 1988; Cucina & Vasilopoulos, 2005).
More recently, Fleeson (e.g., Fleeson, 2001, 2012) has suggested that a personality trait should be thought of as a density distribution of personality states. In other words, the trait is the mean level of trait-relevant behavior across situations when trait-relevant behavior is measured a large number of times. This viewpoint has two notable implications.
The first implication is that there exists considerable variability (inconsistency) in trait-relevant behavior across situations. In Fleeson’s (2001: 1011) own words, “the typical individual regularly and routinely manifest[s] nearly all levels of all traits in his or her everyday behavior.” Fleeson and Gallagher’s (2009) meta-analysis, for instance, demonstrated that 75%, 78%, 62%, 63%, and 49% of the variance in conscientiousness-related, extraversion-related, neuroticism-related, agreeableness-related, and openness-related behavior, respectively, was attributable to differences across occasions (time points) rather than to differences across people. When restricted solely to workplace settings, trait-relevant behavior exhibits slightly lower but nonetheless considerable within-person variability across occasions (Huang & Ryan, 2011; Judge, Simon, Hurst, & Kelley, 2014).
The second implication of Fleeson’s (2001, 2012) work is that this variability in trait-relevant behavior across situations, like the mean level, appears to be relatively stable over periods of time (see also Mischel, Mendoza-Denton, & Shoda, 2002, who studied variability of the “if [situation], then [behavior]” form and concluded that such variability is often stable over periods of time). This suggests that the mean and variability reflect two relatively distinct individual differences. To summarize, then, Fleeson’s work indicates that (what we refer to as) personality strength is meaningful and that individual differences in personality strength accompany individual differences in personality traits.
Findings From Extant Literatures
A broad empirical question germane to the nature and effects of personality strength is this: What sorts of previous studies have examined concepts relevant to our definition of personality strength, and what have these studies found? To address this question, we reviewed extant personality research on constructs that could be viewed anew through the lens of personality strength: for instance, the literatures on traitedness, personality stability, personality consistency, intraindividual consistency, intraindividual predictability, self-monitoring, and hardiness (e.g., Baumeister, 1991; Bem & Allen, 1974; Bornstein & Cecero, 2000; Cucina & Vasilopoulos, 2005; Eschleman, Bowling, & Alarcon, 2010; Gangestad & Snyder, 2000; Paunonen, 1988). We also reviewed the literature on attitude strength (e.g., Cooke & Sheeran, 2004; Krosnick & Petty, 1995) and the literatures on self-related constructs such as self-concept clarity and self-certainty (e.g., Campbell, Trapnell, Heine, Katz, Lavallee, & Lehman, 1996; Kernis, 2003) because several of the ideas and operationalizations from these literatures could readily be translated to the domain of personality strength.
Our review of the aforementioned literatures suggests at least four important conclusions. First, a common prediction involving personality strength appears to have developed, largely independently, across several of these literatures. We discuss this prediction in the next section. Second, although some of these literatures cite foundational work on situational strength (e.g., Mischel, 1968), their predictions related to personality strength are virtually never inspired by the situational strength literature. In the next section, we therefore delineate three novel situational-strength-inspired research predictions that situate personality strength as an important component of person-situation interactions. Third, the overall set of personality strength operationalizations generated through our review of the literature lends itself to a relatively parsimonious three-category structure of personality strength. We discuss this categorization in a subsequent section of the article. Fourth, despite the importance of personality strength in organizational settings, and despite the promise of organizational research methods such as situational judgment tests in operationalizing personality strength, there is a dearth of studies using employees or job applicants (as opposed to undergraduate students) as respondents. We provide several suggestions for the operationalization of personality strength in a subsequent section of the article, with a special emphasis on workplace-relevant operationalizations.
Primary Predictions From a Personality Strength Lens
We use this section to describe four predictions: one prediction in common across many of the literatures we reviewed, and three predictions informed by the situational strength perspective. All four predictions take the form of interactions. The common prediction and the first of the new predictions can be represented as Behavior = f(Person, Person) and Behavior = f(Person, Situation), respectively. In other words, these two predictions are both interactions of the traditional moderation variety. In contrast, the other two new predictions are interactions of the “systems” variety (Buss, 1977; Zuroff, 1986). Both these predictions focus on dynamic Situation = f(Person) relationships. 1
Common Prediction
Many of the operationalizations of personality strength taken or adapted from extant literatures generate a common prediction: namely, that the relationship between a personality trait and behavior—or, in organizational settings, job performance—is moderated by (what we refer to as) personality strength, such that trait–behavior relationships are stronger when personality is stronger. Of note is the fact that similar predictions appear to have been generated in the attitude and self-concept literatures largely independently of each other and the personality literature (Baird, Le, & Lucas, 2006; S. J. Sherman & Fazio, 1983; note also the rarity of cross-literature citations in sources such as Cooke & Sheeran, 2004; Cucina & Vasilopoulos, 2005; Gangestad & Snyder, 2000; Kernis, 2003; Krosnick & Petty, 1995). Thus, uncovering this common prediction is a benefit of collating these literatures under the rubric of personality strength. We note, however, that this common prediction is not directly relevant to (i.e., it neither supports nor falsifies) the theoretical model described in the next subsection and in Figure 1.
Proposed Model of the Impact of Personality Strength on Within-Person Variation in Behavior
Our review indicates that, although prominent unsupportive findings do exist (e.g., Chaplin & Goldberg, 1984; Cucina & Vasilopoulos, 2005), the common prediction has frequently been supported at least in part in extant published research. For instance, Barrick, Parks, and Mount (2005) found that self-monitoring (an operationalization of personality strength) attenuated the relationships between the Big Five personality traits and supervisor reports of the interpersonal aspects of job performance, such that, on average across the Big Five personality traits, the Personality Trait × Self-Monitoring interaction explained an additional 5.68% of variance in interpersonal job performance beyond the 4.22% explained by the main effects of both the personality trait and self-monitoring. As another example, consider the domain of attitude strength (some operationalizations of which can readily be modified to serve as operationalizations of personality strength). Cooke and Sheeran’s (2004) meta-analysis found that, on average across seven operationalizations of attitude strength (e.g., accessibility, affective-cognitive consistency), attitude–behavior correlations were stronger when attitude strength was high (mean r = .53) than when it was low (mean r = .25).
Given the modest main effects for personality traits in predicting job performance (Barrick & Mount, 1991), the main effects of personality strength and the Personality Trait × Personality Strength interactions are likely to be incrementally meaningful. This is especially likely to be the case when best practices regarding the operationalization and measurement of personality strength are adhered to—a topic we discuss in a subsequent section of the article.
Novel Predictions
We turn now to the unique predictions using a personality strength “lens.” These predictions are summarized in Figure 1. Of note is the fact that personality traits have traditionally been used to predict characteristic levels of behavior or performance, and this outcome variable is retained in the aforementioned common prediction. Importantly, however, a personality strength lens would suggest a different outcome: within-person variability in behavior or performance.
Prediction 1: Attenuation of impact of situations
Recall that our definition of personality strength—designed to parallel that of situational strength—emphasizes the extent to which internal cues suggest that potential behaviors are desirable to the individual, irrespective of situational forces. This definition leads directly to the first novel prediction. To the extent that internal cues are emphasized, they compete with external cues in the prediction of behavior. Stated differently, we predict that situations (conceptualized in terms of either content or strength; Meyer et al., 2010) will have a less pronounced effect on the behavior of people with stronger personalities compared to people with weaker personalities. As alluded to previously, the relationship that we propose will be attenuated by personality strength is that between within-person variability in situations and within-person variability in behavior (including job performance). Specifically, we articulate the following proposition (see also Figure 1):
Proposition 1: Personality strength moderates the relationship between within-person variability in perceived situations experienced over time and within-person variability in behavior expressed over time, such that the relationship is weaker for strong personalities than for weak personalities.
Prediction 2: Emphasis on homogenous aspects of situations
In addition to predicting that personality strength will influence people’s behavioral reactivity to situational cues, we predict that possessing a strong personality will influence people’s perceptions of the situations they experience. Specifically, we predict that people with strong personalities will be more likely than their weak personality counterparts to focus on situational cues that encourage behavior consistent with their trait profiles, and that they will be less likely to focus on situational cues that encourage behavior inconsistent with their trait profiles. At work, relevant situational cues might emanate from (among others) supervisors, colleagues, the organizational climate, the organization’s policies and procedures manual, and the nature of the work itself. According to situational strength theory, to the extent that each of these sources conveys consistent information, the situation will be stronger and employees’ behavior will be homogenized (Meyer et al., 2010). The personality strength perspective qualifies this idea by suggesting that another type of consistency also matters for strong personalities: specifically, consistency between situational and personality cues to behavior. People with a strong personality will emphasize the sources of situational information that are consistent with the information provided by their own personality trait profiles, thereby emphasizing situational information that supports their already-preferred course of action and deemphasizing information that contradicts it. Specifically, we articulate the following proposition (see also Figure 1):
Proposition 2: Personality strength is related negatively to within-person variation in perceived situations experienced over time.
Prediction 3: Proactive choice and shaping of situations
The previous prediction (Prediction 2) referred to perceived or subjective situations. For the current prediction (Prediction 3), we turn to objective situations. These are “perceptually unfiltered objective sensory stimuli encountered by individuals” (e.g., temperature) or at least “consensually accepted definition[s]” of sensory stimuli (e.g., organizational, as opposed to psychological, climate; Hattrup & Jackson, 1996: 511). Researchers who study situations (e.g., Block & Block, 1981; Hattrup & Jackson, 1996) generally suggest that subjective situations (the focus of Proposition 2) are a function of both objective situations and personality. However, consistent with the premises of dynamic interactionism (i.e., the notions that people constantly influence situations and vice versa; Zuroff, 1986; see also Buss, 1977), we further propose (in Proposition 3) that personality strength influences the nature of the objective situations people experience. This effect can occur through two mechanisms: situational choice and situational shaping.
First, people make choices about the situations they are willing to enter (Schneider, 1987; Snyder & Gangestad, 1982), and people with strong personalities are likely to make more, and more definitive, choices in this regard than people with weak personalities. Decisions by people with strong personalities regarding whether to enter a situation are, furthermore, likely to be influenced by the extent to which they believe that the behavioral requirements of the situation are consistent with their own behavioral predilections (i.e., personality profile; Allport, 1937). Empirical support for this idea comes from the attitude strength literature, in which researchers have found that individuals with strongly held attitudes are more likely to selectively encounter situations that provide information consistent with their attitudes (Brannon, Tagler, & Eagly, 2007).
Second, people with strong personalities are likely to try to shape the situations they encounter, such that the behavioral requirements of these situations conform to their own behavioral predilections. Such people actively convey information (backed, in some cases, by incentives and disincentives) to others regarding which behaviors are versus are not “appropriate” in that situation. Specifically, we articulate the following proposition (see also Figure 1):
Proposition 3: Personality strength is related negatively to within-person variation in objective situations experienced over time.
Figure 1 combines these propositions into an overall model of the predicted effects of personality strength. In addition, however, the overall model provides suggestions regarding the relative importance of the three propositions. Specifically, the model suggests that the proposition in which interactions take the traditional form of statistical moderation—namely, the attenuation by personality strength of the impact of experienced situations on behavior (Proposition 1)—may primarily be important in the relatively infrequent instances in which individuals suddenly find themselves, by happenstance or coercion, in situations wherein their behavioral proclivities are contrary to the demands of the situation. In contrast, in the presumably more frequent instances in which individuals are able to choose, shape, and subjectively interpret objective situations (i.e., when interactions are of the systems variety; Buss, 1977)—see Propositions 2 and 3—the impact of situations will already be consistent with that of personality strength.
Stated differently, in contrast to the “either personality or situations” approach to Person × Situation interactions adopted by the traditional statistical moderation perspective on interaction, the systems perspective adopts a “both personality and situations” approach. Thus, in cases in which people are able to choose, shape, and subjectively interpret situations—in other words, when “the people make the place” (Schneider, 1987)—the impact of personality strength may already be “baked into the cake” via situations that are consistent with personality strength. As a result, personality strength may not further attenuate the impact of situations on behavior. All in all, then, we suggest that research is likely to find stronger effects of personality strength on the “upstream” processes of situational choice, shaping, and interpretation than on the “downstream” process of attenuation of situational effects on behavior (or job performance).
Relative importance aside, however, the propositions depicted in Figure 1 are important because they permit not only better prediction of behavior—that is, higher criterion-related validity—but also a better psychological explanation of the interrelationships among persons, situations, and behavior. After all, although the main effects of personality and situations are typically larger than their interactive effects, few researchers truly believe that behavior is merely the additive result of personality and situations (Chaplin, 1997). Isolated and partial exceptions aside (e.g., Lippa & Donaldson, 1990; Snyder & Gangestad, 1982), the interactions proposed here have not yet received research attention. Future research, conducted under the banner of personality strength, should therefore test these propositions using multiple operationalizations of personality strength. To that end, the next section of this article categorizes the large number of operationalizations we reviewed into broader categories of operationalizations.
Categories of Personality Strength Operationalization
As indicated previously, operationalizations of what we call personality strength emanate from several disparate literatures. This has led to researchers using different terms for similar constructs and vice versa (i.e., the “jingle” and “jangle” fallacies; see Block, 1996). Thus, one of our primary goals with the present review is to consolidate these disparate perspectives under a single theoretical umbrella—that of personality strength.
A useful way of organizing these literatures is to focus on general categories of personality strength operationalization. Three such categories emerge from our reading of these literatures: (a) statistical operationalizations (i.e., indices of the extent to which personality is consistent across situations and/or stable over time), (b) content-general operationalizations (i.e., individual difference concepts important less for their own content than for their indication of one’s overall responsiveness to situational cues), and (c) content-independent operationalizations (i.e., operationalizations that assess the situational responsiveness of personality traits, but that are not inherently yoked to any specific trait). We provide examples of operationalizations in each of these three categories. In this section of the article, our focus is description, not evaluation. However, in a subsequent section, we return to these operationalizations for a discussion of “best practices” in operationalization and measurement.
Statistical Operationalizations
Within-person standard deviation
Perhaps the simplest way to quantify the extent to which a given person’s behavior changes across situations (or time) is to calculate a within-person standard deviation of trait-relevant behavior. This within-person standard deviation quantifies the “absolute similarity” (Fleeson & Noftle, 2008: 1372) of the person’s behavior from situation to situation. Smaller (narrower) standard deviations represent high personality strength and larger (wider) standard deviations represent low personality strength.
Other statistical operationalizations
The aforementioned within-person standard deviation across situations associated with a focal personality construct (e.g., conscientiousness) can be ipsatized by dividing it by the within-person standard deviation across situations for the personality inventory as a whole (Cucina & Vasilopoulos, 2005). Alternately, the variance can be used in lieu of the standard deviation (e.g., Bem & Allen, 1974). Another alternative is to examine the extent to which personality expression exhibits specific patterns (i.e., profiles) of variability across various types of situations—for instance, high conscientiousness in work situations but low conscientiousness in family situations (Mischel et al., 2002). A somewhat different operationalization involves an “appropriateness” or “person-fit” index (e.g., Lz; Drasgow, Levine, & Williams, 1985), which identifies people whose response patterns across various trait-relevant situations are inconsistent with their estimated trait level (Reise & Waller, 1993; Schmitt, Cortina, & Whitney, 1993; see also Lanning, 1988): for example, job applicants who exhibit some responses consistent with high conscientiousness but who nonetheless end up being adjudged very low overall in conscientiousness.
Content-General Operationalizations
Throughout the history of personality psychology, several individual difference variables have pertained to the extent to which one’s behavior is influenced by situational cues. These conceptualizations are content-general because the assumption is that behavioral responsiveness to situations is a key human characteristic in and of itself that can be manifested in ways that are not yoked to any specific type of behavior or any broad domain of situations. In other words, these variables have typically been prized more for their effects than for their own content. It is important to note that unlike the statistical and content-independent operationalizations of personality strength, which are trait-specific, the content-general operationalizations reflect a more parsimonious view of personality strength. Stated differently, the content-general perspective suggests that a single personality strength construct can be applied to all personality traits. Here, we focus primarily on self-monitoring but also briefly highlight a few other individual differences.
Self-monitoring
Individuals who are willing and able to project a situationally appropriate façade are referred to as high self-monitors, whereas those who are unwilling or unable to adapt their behavior to the press of the situation are referred to as low self-monitors (Gangestad & Snyder, 2000; Snyder, 1974). High self-monitors may expend considerable time and energy to prepare for social interactions, may gravitate toward situations that are predictable and/or well defined, may experience high levels of self-consciousness during social interactions, and may rely on the behavior of others as an indicator of socially appropriate actions (Ickes, Holloway, Stinson, & Hoodenpyle, 2006). Self-monitoring is therefore a suitable operationalization of personality strength, such that a high score on a measure of self-monitoring reflects a weak personality whereas a low score reflects a strong personality.
A meta-analysis of the role of self-monitoring in organizational settings (Day, Schleicher, Unckless, & Hiller, 2002) yielded a main effect of self-monitoring on job performance: meta-analytic r = .10 after correcting for unreliability in the measure of self-monitoring. Importantly, however, our focus is quite different from that of Day et al. We discuss behavior or performance variability as an outcome of personality strength—a focus consistent with self-monitoring theory (Gangestad & Snyder, 2000)—and, in so doing, aim to explicate interactions between personality strength and both objective and subjective situations.
Additional content-general operationalizations
Several individual difference variables have been used in research on the effects of environments on behavior to help account for individual differences in the extent to which people’s behaviors are influenced by situational cues. For example, consider interpersonal dependency (Bornstein & Cecero, 2000). High scorers exhibit a need for guidance and approval from others (akin to the weak personality), whereas low scorers are more autonomous (akin to the strong personality). As another example, the concept of hardiness (Eschleman et al., 2010; Funk, 1992; for related concepts, see also Duckworth & Quinn, 2009; Youssef & Luthans, 2007), which we discuss in more detail subsequently, has been used as an explanatory mechanism for why certain individuals are able to emerge from negative situations relatively unscathed.
Content-Independent Operationalizations
Here, the focus is once again not on directly assessing the within-person variability of trait expression across time and situations, but instead on understanding its likely antecedents. Several content-independent operationalizations can readily be adapted to the personality domain from extant research on constructs such as attitude strength and the strength of the self. Although these operationalizations were originally developed in the context of specific constructs or categories of constructs (e.g., attitudes, self-concept), they can readily be applied to other constructs—hence the reason they can be imported to the domain of personality for use with specific personality traits, and hence the title “content-independent” (and the difference between this category and the “content-general” category).
Operationalizations derived from attitude strength
Analogous to our conceptualization of strong personality, a strong attitude refers to an attitude that is highly relevant to the individual and/or is predicted to remain stable over time and attempts at persuasion (Franc, 1999; Holland, Verplanken, & Van Knippenberg, 2002). Despite a large and thriving body of research on job-related attitudes (Dalal, 2013), research on attitude strength has been conducted primarily by social psychologists. This extant research suggests that attitude strength should moderate the attitude–behavior relationship, such that this relationship is stronger when attitude strength is high (Krosnick & Petty, 1995; Pomerantz, Chaiken, & Tordesillas, 1995). Empirical results have mostly been supportive (e.g., Cooke & Sheeran, 2004; Franc, 1999). In this sense, extant research on attitude strength, when generalized to personality, contributes to the “common prediction” discussed previously.
Attitude strength has been operationalized in several ways (Costarelli & Colloca, 2007; Krosnick & Petty, 1995; Pomerantz et al., 1995): as accessibility (the speed with which a person responds to questions regarding an attitude object), ambivalence (the extent to which a person tends to simultaneously evaluate an attitude object strongly positively and strongly negatively), certainty (the extent to which a person is confident of the correctness of his or her attitude toward an object), ego involvement or centrality (the relevance or centrality of an attitude toward an object to a person’s self-concept), importance (the importance assigned by a person to an attitude object), and knowledge (the amount of information a person possesses about an attitude object), to take just a few examples. Some of these operationalizations are typically measured via self-report (e.g., importance), others via either objective tests or self-report (e.g., knowledge), and yet others via formulae (e.g., ambivalence). Moreover, several of these operationalizations can readily be adapted to personality. Vis-à-vis extraversion, for instance, importance could refer to the importance assigned by a person to extraverted behavior (see, e.g., Paunonen, 1988) whereas accessibility could refer to the speed with which a person responds to items in an extraversion scale.
Other operationalizations
Similar to the manner in which researchers have operationalized the strength of attitudes in several ways, other researchers have operationalized what might be termed the strength of the self in several ways: for example, self-view certainty (the extent to which a person is confident regarding specific beliefs about the self; Anseel & Lievens, 2006; Baumgardner, 1990), self-concept clarity (the extent to which a person’s self-beliefs are clearly defined, internally consistent, and temporally stable; Campbell et al., 1996), self-esteem stability (the extent to which people’s “self-concepts lack internal consistency and temporal stability and are held with little confidence”; Kernis, 2003: 2), and the presence of self-schemata or self-scripts (“cognitive generalizations about the self, derived from past experience, that organize and guide the processing of self-related information contained in the individual’s social experiences”; Markus, 1977: 64). Conceptually, such developments seem particularly faithful to the notion, espoused by early theorists (e.g., Rogers, 1961), that within-person variability is the result of a weak sense of self (Baird et al., 2006). These operationalizations, like those of attitude strength, can readily be modified to focus on personality.
Potential Relationships Among the Three Categories of Operationalization
An important question for future research pertains to the likely interrelationships among the statistical, content-general, and content-independent operationalizations discussed in the previous section. To this end, we provide two possibilities.
One possibility, suggested by the empirical relationships observed among personality strength operationalizations (e.g., Eid & Diener, 1999; Paunonen, 1988; though see Reise & Waller, 1993), is a hierarchical structure with individual operationalizations of personality strength at the base, categories of operationalization (not necessarily limited to the three discussed above) as first-order factors, and a single factor (or very small number of factors) at the apex denoting the personality strength associated with a particular personality trait.
However, there exist alternatives to this hierarchical structure. Here, we focus on one such alternative, according to which the three categories of operationalization exhibit causal interrelationships. In this regard, it has been argued previously (e.g., Zuroff, 1986) that different personality traits may be thought of in terms of different goals of scientific research: description (merely documenting the nature of the phenomenon), prediction (forecasting the conditions under which the phenomenon will occur), and explanation (elucidating why the phenomenon occurs). We contend that personality strength, too, can be thought of in terms of these goals. Statistical operationalizations may be viewed as describing personality strength. People whose personality remains consistent across situations are, ipso facto, those who possess strong personalities. Content-general operationalizations may be viewed as predicting personality strength. People who score low on self-monitoring, for instance, are those whose standing on a personality trait is likely to remain consistent across situations (Britt, 1993). Content-independent operationalizations may be viewed as explaining personality strength. People’s standing on a personality trait may remain consistent across situations because (for instance) that personality trait is particularly important to them (see, e.g., Britt, 1993).
In this alternative structure, content-independent operationalizations may mediate the relationship between content-general operationalizations and statistical operationalizations. This alternative structure moreover suggests a trade-off between scientific understanding and empirical support. Specifically, because explanation is the highest scientific goal, content-independent operationalizations may best facilitate scientific understanding. Yet because statistical operationalizations are the most proximal manifestations of personality strength, they may best facilitate empirical support for theoretical propositions. Future research should test these ideas.
Recommendations for Operationalization and Measurement
We hope the aforementioned sections provide readers with useful insights into personality strength’s structure and effects. One issue, however, is that tests of both structure and effects require adequate sample sizes. With regard to structure, future studies employing covariance structure analyses should include at least 10 respondents per estimated parameter (Aguinis & Harden, 2009). With regard to effects, the number of respondents needed to test a moderation hypothesis in a hierarchical multiple regression analysis—assuming the archetypal small effect size f2 = 0.02 (because effect sizes in statistical moderation are often small; Chaplin, 1997) and a typical Type I error rate of 0.05—is 395 for an a priori power of 0.80 and 528 for an a priori power of 0.90 (Faul, Erdfelder, Buchner, & Lang, 2009). The extant studies we reviewed that tested the common prediction regarding the Personality × Personality Strength interaction typically had sample sizes considerably smaller than these guidelines, making interpretation difficult in cases where these studies failed to find hypothesized effects. Future studies testing this hypothesis as well as our Proposition 1 (which also involves statistical moderation) should therefore use sample sizes considerably larger than those currently used.
Yet empirical findings regarding both structure and effects are dependent on much more than just adequate statistical power. They also require optimal operationalization of the focal construct. We now focus on several important issues that we hope will guide future operationalization and measurement in each of the three categories of operationalization described previously. Table 1 provides a set of “best-practice” recommendations.
Best Practices in the Operationalization of Personality Strength: Suggestions for Future Research
Statistical Operationalizations
Personality trait versus personality strength
The within-person standard deviation of personality trait expression (an operationalization of personality strength) is not independent of the within-person mean (an operationalization of personality trait). Theoretically, the mean and standard deviation exhibit an inverse U-shaped relationship, such that the standard deviation is highest when the mean is at the midpoint of the scale and lowest (zero) when the mean is at the endpoints of the scale. When using the standard deviation as an operationalization of personality strength, therefore, researchers should control for both the mean score and the square of the mean score (Baird et al., 2006; Eid & Diener, 1999).
Sampling across situations
A potential complicating factor exists with respect to the psychological meaning of the statistical operationalizations as metrics of personality strength. Specifically, it is assumed that the sample of situations utilized is a representative sample of the population of situations that people might experience. If, however, the situations used to calculate the within-person standard deviation are objectively or subjectively similar, the proportion of variance that is attributable to the person will be inflated. 2
Perhaps the best way to ensure an adequate range of situations involves using a situational taxonomy. Unfortunately, because there is considerably less consensus about the structure of situations than the structure of personality (Funder, 2006; Johns, 2006), researchers may simply have to choose a particular taxonomy of situations and justify its use. Space considerations here preclude a listing of the many candidate taxonomies and their pros and cons, but researchers will need to choose between a general taxonomy of situations (e.g., Yang, Read, & Miller, 2006) and a taxonomy of trait-relevant situations (e.g., Rauthmann et al., 2014; Ten Berge & De Raad, 2002; Wagerman & Funder, 2009). Organizational researchers will moreover need to choose between domain-general and workplace-specific situational taxonomies (e.g., Barrick, Mount, & Li, 2013; Humphrey, Nahrgang, & Morgeson, 2007).
Regardless of which situational taxonomy is chosen, one then has to decide how to sample situations from that taxonomy. In an ideal world, one would sample randomly from the multivariate distribution of situations within the chosen taxonomy (e.g., a quadrivariate normal distribution for Ten Berge & De Raad, 2002). A less optimal but considerably simpler approach would consider each situational dimension individually and would ensure adequate representation of situations high, moderate, and low on each dimension.
Pros and cons of various research designs
The need to sample an adequate range of situations will play out differently in various research designs. Experience sampling methods (Hektner, Schmidt, & Csikszentmihalyi, 2007) involve measurements on a large number of occasions (time points) per person. This provides the potential for obtaining measurements in a wide variety of situations, but the equivalence of occasions and situations is unlikely because a person may experience the same situation on multiple occasions and may not experience other situations at all (Diener & Larsen, 1984; Fleeson & Noftle, 2008). 3 As such, researchers interested in assessing personality strength via experience sampling studies should attempt to maximize situational variance not only by increasing the number of occasions on which each person is surveyed (which would have the added benefit of increasing the reliability of statistical operationalizations of personality strength; Baird et al., 2006) but also by exploiting features of their research designs that increase situational variance.
For example, by deliberately surveying respondents at multiple times of the day and on multiple days of the week, researchers would be able to assess behavior enacted in multiple life domains (e.g., work vs. home) and during various levels of activity (e.g., restless vs. sluggish; Watson, 2000). Researchers could also ask participants to respond to surveys upon encountering specific types of situations (e.g., those that would otherwise be undersampled). These options—an “interval-contingent” design and an “event-contingent” design, respectively (Beal & Weiss, 2003)—are not mutually exclusive: Researchers could overlay the latter on the former.
A different possibility is to examine behavior in the context of an assessment center (Gaugler, Rosenthal, Thornton, & Bentson, 1987). An assessment center is characterized by several “exercises”—for instance, a leaderless group discussion, an in-basket exercise, and a role-play—but exercises should not be confused with situations: Multiple exercises may share the same underlying situational characteristics (Gibbons & Rupp, 2009). Moreover, the number of exercises assessed in an assessment center (on average, 7; Gibbons & Rupp, 2009) is much too small for an adequate range of situations to be assessed. Ideally, then, exercises would reflect “pure” situational types (as determined by a situational taxonomy) and assessment centers would include a considerably larger number of exercises. The former is achievable, but, given time and cost constraints, the latter may prove prohibitive.
These limitations are unfortunate, given that assessment centers have some advantages over experience sampling designs. One such advantage involves the fact that assessment center exercises could be defined as objective rather than subjective situations if subject-matter experts agree as to their situational ingredients (e.g., Hattrup & Jackson, 1996). Asking respondents about the extent to which they would choose to “enter” various exercises of their own free will would permit tests of our prediction that personality strength is related negatively to within-person variation in objective situations (Proposition 3). Another advantage that assessment centers have over experience sampling designs is that the range of situations in experience sampling designs must be measured (and subsequently controlled for statistically) whereas it can be manipulated in assessment centers—thereby permitting stronger causal conclusions.
An alternative design that preserves the advantages of assessment centers while allowing for greater tractability vis-à-vis its disadvantages is the situational judgment test (McDaniel, Morgeson, Finnegan, Campion, & Braverman, 2001). Like the exercises in an assessment center, the vignettes in a situational judgment test could be defined as objective (as opposed to subjective) situations if subject-matter experts agree as to their situational ingredients. Moreover, like an assessment center, a situational judgment test can involve situations that represent pure situational types and situations that are manipulated as opposed to merely being measured. Yet because vignettes in a situational judgment test are typically “simpler” (less elaborate, easier/quicker to explain to participants, etc.) than exercises in an assessment center, using a sufficient number of situations is more readily achievable in a situational judgment test than in an assessment center.
To summarize, each of the aforementioned research designs has its pros and cons—but even the best of these designs must be modified so that respondents’ scores on the statistical operationalizations are good reflections of their personality strength. We contend that support for the three propositions we have previously outlined (see Figure 1) will be stronger when scores on personality strength operationalizations are assessed via research designs optimized in the manner discussed above. Future research should assess this contention.
Prior to closing this subsection, however, we note that most existing research on cross-situational consistency, traitedness, and so forth has not used any of the aforementioned research designs. Rather, this research has typically used the within-person standard deviation (or alternative statistical operationalization) of scores across items in a traditional personality inventory administered on a single occasion. Even aside from the fact that responses to an individual personality item are in part attributable to item-specific factor error and random response error (Putka, Fleisher, & Beatty, 2014; Schmidt, Le, & Ilies, 2003), measuring personality strength in this way would be viable only if the personality inventory were completely reinvented. Each item in the inventory would have to be written to reflect a pure situational type, and the number of items would have to be sufficient to represent an adequate range of situations. 4
Variability in personality versus variability in behavior
Finally, an important issue is how to distinguish within-person variability in a personality trait from within-person variability in trait-relevant behavior (including various forms of job performance; Dalal, Bhave, & Fiset, 2014; Dalal, Lam, Weiss, Welch, & Hulin, 2009). This is a specific instantiation of a broader distinction—between personality and behavior—that has bedeviled the field of personality psychology since nearly its inception. Philosophical treatments of personality traits differ regarding whether they view traits as causes of behavior (with traits themselves being viewed as genetically determined through as yet largely unspecified mechanisms) or simply as summaries of past behavior that provide a basis for the estimation of future behavior (Buss & Craik, 1983). The latter perspective does not distinguish personality from behavior, and arguably the former perspective does not either (though it does distinguish biological substrates from behavior).
Given these long-standing difficulties in conceptually distinguishing personality from behavior, it would be a tall order for us to do so definitively in these pages. Nonetheless, as Wagerman and Funder (2009) noted, researchers can study personality, situations, and behavior successfully only if these constructs are conceptualized and measured independently—the alternative leads ultimately to theoretical paralysis. Moreover, the within-person level, with its focus on personality states rather than traits, brings the issue into even starker relief. Is it possible to measure personality in such a way that within-person fluctuations in personality are not redundant with within-person fluctuations in behavior during the same time period? Stated differently, within-person variation arguably makes the case for “personality qua personality” (as distinct from “personality qua behavior”) all the more urgent. The propositions we have outlined in the current article are dependent on such a distinction being feasible.
We therefore offer two perspectives, neither of which is completely satisfactory. The first perspective is to define personality as a set of behavioral predilections or tendencies. We view these as distal, dispositionally driven behavioral preferences that, in conjunction with situational strength, determine behavioral intentions. Because behavioral predilections can vary across situations, and because behavioral predilections do not always get translated into actual behavior (in much the same way that behavioral intentions do not; Webb & Sheeran, 2006), at the within-person level personality and behavior can both vary without being redundant.
The second perspective relies on the possibility of measuring personality states not just as behavior (e.g., “How hard are you working on this task right now?”) but also as cognition (e.g., “How mentally focused are you on this task right now?”) and affect (e.g., “How anxious does this task make you feel right now?”; see, e.g., Beckmann, Wood, & Minbashian, 2010). If, as has been suggested for attitudes (Dalal, 2013; Dalal, Brummel, Wee, & Thomas, 2008), personality is conceptualized only as cognition and emotion, then personality and behavior could vary nonredundantly within persons. This perspective is by no means mutually exclusive with the first perspective because cognition and emotion together constitute a behavioral predilection.
Content-General Operationalizations
Are they truly content-general?
Previously, in the context of statistical operationalizations of personality strength, we had suggested that researchers measuring personality strength should control statistically for both the personality trait score and the square of the personality trait score. Our rationale there was statistical. A similar issue arises in the case of content-general operationalizations, however, and this time our rationale is conceptual. At issue here is whether these operationalizations are truly content-general.
For instance, research on self-monitoring has repeatedly had to assert the distinctiveness of self-monitoring from the personality trait of extraversion (Gangestad & Snyder, 2000; Lennox & Wolfe, 1984). Moreover, some researchers (e.g., Britt, 1993) have used arguments similar to those in the self-monitoring literature to implicate extraversion itself as an operationalization of what we would refer to as personality strength. Therefore, researchers using self-monitoring as an operationalization of personality strength could examine whether their findings change when they control statistically for extraversion. In fact, because of the possibility of nonlinear relationships between personality traits and personality strength (Cucina & Vasilopoulos, 2005), we suggest controlling for both extraversion and squared extraversion scores.
As another example, consider interpersonal dependency and hardiness. Both these constructs are related empirically to the personality trait of neuroticism (Bornstein & Cecero, 2000; Funk, 1992). Moreover, some researchers (e.g., Judge et al., 2014; R. A. Sherman, Nave, & Funder, 2010; Snyder & Monson, 1975) have implicated neuroticism itself as an operationalization of what we would refer to as personality strength. Therefore, researchers using interpersonal dependency and hardiness as operationalizations of personality strength could control statistically for both neuroticism and squared neuroticism scores.
Trait-specific or individual difference concepts?
Content-general operationalizations of personality strength, for example self-monitoring, are typically conceptualized as individual differences. In contrast, statistical and content-independent operationalizations can differ across personality traits: These operationalizations permit a person to have a strong personality for (say) conscientiousness that coexists with a weak personality for (say) neuroticism. The lack of trait-specific differentiation may therefore be a limitation of content-general operationalizations. Alternatively, however, personality strength may be a single individual difference variable that stems from a weak sense of self (Baird et al., 2006; Rogers, 1961; but see in contrast Kenrick & Stringfield, 1980) and that is equally relevant to all personality traits. A middle-ground perspective would hold that personality strength consists of both individual difference and trait-specific components (Baumeister, 1991).
The manner in which this issue is ultimately resolved has important implications for parsimony in personality research and its application in organizational settings. Consider, for instance, the Big Five personality traits. If personality strength is a content-general individual difference variable, a single personality strength construct would apply to all five personality traits. In contrast, if personality strength is trait-specific, five personality strength constructs would be needed—one for each trait.
One-sided (or local) operationalizations
Fleeson and Noftle (2008) note that trait-relevant behavior may, in some cases, be locally rather than globally consistent. In other words, a person could exhibit high personality strength across only a portion of the situational continuum rather than across the whole continuum. An example is provided by the construct of hardiness, which (along with related constructs such as resilience and grit) refers to individuals’ ability to perceive difficult, or, more broadly, negative situations as engaging (rather than alienating) and challenging (rather than threatening) and to feel in control (rather than helpless; Hull, Van Treuren, & Virnelli, 1987). This focus on negative situations suggests that the reactivity of hardy people to positive situations is undiminished. As such, hardiness may be viewed as a “one-sided” (asymmetrical) operationalization of personality strength—and perhaps as a subset of stoicism, which involves low reactivity regardless of the valence of the situation (Irvine, 2009). An operationalization that is one-sided (or, more generally, local) may nonetheless have important practical implications. For instance, the military has an interest in dampening soldiers’ reactivity to negative situations (e.g., Duckworth & Quinn, 2009) but presumably cares little about their reactivity to positive situations. Future research could therefore focus on understanding whether various operationalizations of personality strength are local or global, and on developing additional local operationalizations.
Content-Independent Operationalizations
When adapting operationalizations from the attitude and self domains to the personality domain, it is important to understand the interrelationships among these constructs. Attitude strength researchers have proposed conceptual groupings of attitude strength operationalizations (e.g., aspects of the cognitive structure of an attitude vs. subjective impressions of the attitude and attitude object), but extant empirical results have generally not supported the view of attitude strength as a single dimension (Eaton, Majka, & Visser, 2008; Krosnick & Petty, 1995).
It is of course important for future research to examine the extent to which these results generalize to personality strength. However, as we discuss in this section, several methodological practices have made it difficult to accurately ascertain the structure of attitude strength. Thus, although our focus is personality strength, our suggestions are equally applicable to attitude strength.
Number of items per operationalization
The first issue is the number of items used to measure each content-independent operationalization of personality strength. Although the desire to minimize survey length is understandable, the use of single-item measures of each operationalization (e.g., Bassili, 1996) is undesirable because a single-item measure does not adequately cover the construct space of an operationalization or its relationships with other operationalizations (let alone its relationships with behavioral criteria).
Use of formulae involving difference scores
The second issue involves the use of formulae to measure operationalizations like ambivalence and affective-cognitive consistency (Costarelli & Colloca, 2007; Franc, 1999). Formulae such as these contain all the psychometric problems of difference scores: for example, confounded effects, low reliability, and untested constraints (Edwards, 2002). Researchers should instead employ polynomial regression and response surface analysis (Edwards, 2002). These techniques do not provide a single score for the congruence (i.e., consistency or nonambivalence) between the two components of the operationalization. Rather, the response surface is based on lines of congruence and incongruence, which provides a deeper understanding of the phenomenon.
Conflating strength with method
An assessment of the structure of personality strength is hindered if different operationalizations are assessed via different methods. For instance, personality accessibility may be measured via response latency, whereas personality relevance may be measured via self-report. Indeed, in some cases, a particular operationalization may be measured using multiple methods—for instance, personality knowledge may be measured both objectively and subjectively—and relationships across methods are often not high. In an ideal world, researchers would be able to use multitrait multimethod (or rather multistrength-operationalization multimethod) confirmatory factor analytic approaches to assess the structure of personality strength. However, because most operationalizations can be measured only via a single method, perhaps the best that can be done is to fit a single method factor (in addition to substantive factors) associated with the use of non-self-report methods.
Practical Implications
A foolish consistency is the hobgoblin of little minds. . . .
One of the primary practical implications associated with personality strength pertains to how strong and weak personalities are perceived by other people around them. On the one hand, strong personalities may be viewed as reliable and weak personalities as erratic or as having a weak sense of self (Baird et al., 2006). Consistency, in other words, may be viewed as a virtue when the alternative is unreliability or a lack of character. On the other hand, strong personalities may be viewed as dogmatic and weak personalities as adaptable. Consistency, in other words, may be viewed as a vice when it impedes cognitive and behavioral flexibility. This would be the case if, for instance, hardy individuals are so stubbornly fixated on managing an adverse situation that they ignore the benefits of radically changing the situation or even leaving the situation altogether. Consequently, when designing employee selection systems, organizations should not view personality strength as an inherently “good” or “bad” characteristic of job applicants in the same way as traditional personality traits (e.g., conscientiousness, neuroticism) are often viewed as good or bad.
Rather, with regard to employee selection, personality strength can be used for two purposes. The first purpose is to improve the predictability of job performance via personality traits. Specifically, personality traits are likely to predict the job performance of strong personalities to a greater extent than that of weak personalities. The second purpose is to better understand person–job fit. For instance, strong personalities may be preferred when there is a need to resist strong situations (including stressful situations; Duckworth & Quinn, 2009). In contrast, weak personalities may be preferred when there is a need to resist overconfidence and to display behavioral flexibility. For example, because they tend to learn from feedback and understand the complexities engendered by situational variability, weak personalities (“foxes”) are likely to be more accurate forecasters than strong personalities (“hedgehogs”; Silver, 2012).
Moving beyond a strictly selection-based perspective, organizations could use personality strength to identify the job incumbents most likely to benefit from interventions such as organizational development and job redesign. Strong personalities may be especially responsive to efforts that provide employees the opportunity to choose and shape (i.e., “craft”; Wrzesniewski & Dutton, 2001) their job tasks, such that these tasks cohere into a whole that is consistent with their personalities. In contrast, weak personalities may be more responsive to externally imposed structuring interventions (e.g., assigned goals) aimed at creating strong situations that constrain behavioral variability (e.g., performance troughs; Lee & Dalal, 2011). Finally, organizations could incorporate personality strength into the feedback process. Gaining insight into their “one-size-fits-all” tendency may benefit strong personalities whereas gaining insight into their unique “if . . . then . . .” situation–behavior profiles (Mischel et al., 2002) may benefit weak personalities (Gibbons & Rupp, 2009).
Conclusion
The current article aimed to convince organizational researchers and practitioners that our field’s trait-centric view of personality is limited because it ignores the consistency, or lack thereof, with which people behave across situations. To that end, we reviewed several literatures associated with personality, attitudes, and the self that provided insight into how personality strength could—and should (in terms of “best practices”)—be operationalized and measured.
By doing so, we hope we have set the stage for future research on personality strength. To be sure, this includes research on the operationalization and structure of personality strength, the new propositions summarized in Figure 1, and the proposition common to many of the literatures we reviewed. However, research on personality strength topics not covered appreciably in the current article (e.g., reverse-direction effects of situations and behavior/performance on personality strength, connections between personality variability over short periods such as weeks and long periods such as decades, and connections between personality variability across situations within a person and personality variability across persons within a team) should also benefit from the conceptual grounding and recommendations provided in these pages. The study of situational strength in organizational settings is proceeding apace (e.g., Dalal & Meyer, 2012; Meyer et al., 2010; Meyer et al., 2014), and we hope that the current article provides a similar impetus to “the other side of the coin,” personality strength.