Scales of impression management—also called validity, social desirability, or lie scales—have been an integral part of personality assessment since its early days (see Uziel, 2010b, for a review). Researchers administering personality questionnaires quickly realized that the validity of self-reports is often compromised because (some) respondents bias their responses (Ellis, 1946). Soon after, attempts to control for response biases gained momentum. The use of social desirability scales was popularized following work by Edwards (1953, 1957) who devised an individual difference tool constructed from items of the Minnesota Multiphasic Personality Inventory (MMPI). Following this, more than a dozen scales were created to measure social desirability (Wiggins, 1964), with the most popular one being the Marlowe-Crowne Social Desirability Scale (MCSDS; Crowne and Marlowe, 1960, Crowne and Marlowe, 1964).

Additional developments uncovered limitations of the MCSDS and indicated that social desirability is a two-dimensional construct. Specifically, following distinctions made by Wiggins (1964) and Sackeim and Gur (1978), Paulhus (1984, 2002) described a model for social desirability that differentiates between impression management (IM) and self-deception enhancement (SDE) and introduced the Balanced Inventory of Desirable Responding (BIDR). IM is said to reflect a habitual and conscious tendency to portray an overly positive self-portrait, whereas SDE reflects an unconscious self-serving bias. It has generally been recommended to control for IM in scale development and administration because biases associated with SDE persist beyond the setting of responding to questionnaires (Paulhus, 1984).

The application of social desirability (specifically, IM) scales to control for response bias has become ubiquitous in research and practice across diverse fields, such as personality, social, clinical, and organizational psychology (e.g., Goffin and Christiansen, 2003; Logan et al., 2008; Uziel, 2010b; Perinelli and Gremigni, 2016; Barry et al., 2017; Maltby et al., 2019). Moreover, individuals with high scores on IM scales were considered to experience enduring psychological maladjustment. Specifically, Crowne and Marlowe (1964; see also Crowne 1979) have suggested that a high IM score reflects a defense mechanism that shields individuals with vulnerable (i.e., low and insecure) self-esteem. This vulnerability is associated with social anxiety, defensive and awkward social manners, and poor adjustment (see also Weinberger et al., 1979). Thus, interest in IM scales stemmed from the dire real-life implications that have been attributed to high scorers in addition to their role as validity tools.

Over the years, evidence has accumulated suggesting that IM scales are ineffective as measures of response bias and that they contain more (personality) substance than (response) style (e.g., McCrae and Costa, 1983; Ones et al., 1996; Uziel, 2010b; MacCann et al., 2012; Müller and Moshagen, 2019).¹ Evidence along this line has shown, for example, that IM scales fail to moderate or suppress criterion-related validities of personality traits in occupational contexts (Li and Bagger, 2006), that they fail to moderate self-other agreement on personality traits (Borkenau and Ostendorf, 1992; Uziel, 2014; Müller and Moshagen, 2019), that knowledgeable others attribute reliability and trustworthiness to individuals with high IM scores (De Vries et al., 2014), as do direct behavioral tests of cheating (Zettler et al., 2015). More so, like other personality traits—but unlike a contextual response set—scores on the IM scale show consistency across situations and over time (Lönnqvist et al., 2007), and they reliably predict real-life behaviors (Uziel, 2010b).

However, a persistent question concerns the substance that IM scales measure. Various studies have identified different correlates of IM. For example, McCrae and Costa (1983) have noted that IM is associated with emotional stability and warmth, and Ones et al. (1996) have concluded (in a meta-analysis) that IM is associated with emotional stability, agreeableness, and conscientiousness. More recent approaches have refined the associations, pinpointing elements of interpersonal orientation and social responsiveness. Bou Malham and Saucier (2016) have associated IM with cultural normativity, representing both knowledge and successful application of cultural values by high-IM individuals (see also Elliot et al., 2018). De Vries et al. (2014) have associated IM with honesty-humility, suggesting that high scorers are sincere, fair, and modest (see also Paulhus, 2002).

These approaches are informative in highlighting the correlates of IM. However, there is a need to accommodate their diverse findings under an overarching model. More so, by and large, they are based on cross-sectional data. Arguably, an exhaustive account of IM and its relation to social behavior would incorporate the social context and the dynamics in the expression of this trait in varying settings (Mischel and Shoda, 1995; Uziel, 2010b).

Concerns about social desirability are an integral part of psychological measurement, particularly in personality assessment (e.g., Barry et al., 2017; Müller and Moshagen, 2019; Williams et al., 2019). Scale development and application are often accompanied by control and correction for social desirability (e.g., MacCann et al., 2012; Maltby et al., 2019; Tse et al., 2020), and individuals scoring high on IM scales were considered to hold maladaptive qualities (e.g., Crowne, 1979).

However, the question of the validity of validity scales has yet to be resolved. Cumulative evidence indicates that these scales fail to measure faking, but, instead, tap personality substance (e.g., McCrae and Costa, 1983; Ones et al., 1996; Uziel, 2010b, 2014; De Vries et al., 2014; Zettler et al., 2015; Müller and Moshagen, 2019). However, much of this evidence builds on associations with broad personality constructs (e.g., McCrae and Costa, 1983; Müller and Moshagen, 2019). Recent works have begun to refine our understanding of the qualities that define IM, emphasizing sensitivity to social contexts and a response comprised of an increase in self-control (Uziel, 2010a, 2014). This approach has been supported in controlled experimental settings that manipulated the social context and explored the behavioral responses of individuals along the IM dimension (e.g., Kline et al., 2002; Uziel, 2010a; Uziel and Baumeister, 2012). However, evidence building on extended real-life dynamics in participants’ natural social environment was still lacking, as well as systematic comparisons between IM and related constructs.

To address these questions, the present research utilized an intensive longitudinal design (Bolger et al., 2003), whereby momentary levels of self-control were monitored more than 4,250 times (across 154 participants) alongside participants’ social condition. Results showed that, in general, people reported higher self-control in the presence of others (vs. alone) and that all four traits (IM, SDE, TSC, Agreeableness) were positively associated with episodic self-control. Addressing our main prediction, IM interacted with social condition, such that higher IM was associated with greater episodic self-control in the presence of other people than alone. High-IM individuals reacted to social presence with an increase in their self-control. In what follows, we discuss the implications of the findings.

The first effect that emerged was that in social presence (vs. alone), people reported greater self-control. This finding is consistent with existing studies using a similar experience sampling approach (Hofmann et al., 2012), and with laboratory-based experimental research (Erber et al., 1996). It reflects the notion that our behavior in social contexts is more controlled than our behavior alone because social presence motivates us to narrow discrepancies from desired standards (Carver and Scheier, 1981). Past (Uziel, 2007) and the present findings have shown that there are individual differences in how strong this effect is.

Turning to the main finding, past research has documented high self-control behavior in social contexts among high-IM individuals using behavioral measures (Blackhart et al., 2007; Uziel, 2010a; Uziel and Baumeister, 2012; Guo et al., 2019) and informant reports (Uziel, 2014). The present findings extend these findings to expressions of social behavior during everyday life. According to the IOSC model (Uziel, 2010b), individual differences in IM reflect a predisposition to react to social presence with an increase in self-control to maximize adaptation. High-IM individuals attribute importance to their social image but their behavior, personality correlates, and life outcomes indicate that their goal is not to impress others in deceitful or narcissistic manners but to integrate well into existing social structures (Uziel, 2010b; Bou Malham and Saucier, 2016). Exerting self-control in social contexts is central to the success of their approach because it assists in curbing the expression of impulses and adjusting behavior to external demands (Baumeister, 1982). Given that self-control resources are limited (Baumeister et al., 1998), it makes sense to conserve and apply them strategically when most needed (e.g., in public social contexts). Indeed, greater self-control in social contexts brings them desirable long-term payoffs by, for example, being considered trustworthy and conscientious (Paulhus, 1998; De Vries et al., 2014; Uziel, 2014).

Although the IOSC model is focused on adult personality, high-IM individuals’ behavioral profile (as expressed in the present study) shares commonalities with profiles described in developmental theories, such as in the ‘Adaptive Calibration Model’ (ACM)—an evolutionary-developmental theory of individual differences in the functioning of the stress system (Del Giudice et al., 2011). According to the ACM, adults’ high IM profile relates to a ‘sensitive pattern’ of responsivity, which develops in infancy in the context of attentive maternal care. This profile supports the development of an emotionally stable and socially competent adult personality, characterized by sensitivity to social feedback and the ability to mobilize metabolic and psychological resources when needed. Although more research is needed to study the developmental trajectories associated with IM, the similarity in repose patterns speaks to the potential adaptive benefits of this profile.

Responsiveness to social contexts (with an increase in self-control) was uniquely observed for IM as compared with the other traits measured. Although self-control is not the central component in the self-perception of individuals high on SDE (Paulhus, 1984), previous studies have found that high-SDE individuals report high self-control. However, informants did not corroborate these attributes (Uziel, 2014). Therefore, high-SDE individuals’ report of greater self-control (alone and in the presence of others to the same extent) is consistent with their self-deceiving bias and low sensitivity to social contexts (Paulhus, 1984; Uziel and Cohen, 2020).

TSC was also associated with greater self-control alone and with other people, as expected based on the definition of this trait (Tangney et al., 2004). Interestingly, recent approaches to TSC have suggested that high-TSC individuals do not necessarily exert more self-control on a momentary basis, because they strategically navigate their environment to reduce self-control conflicts (De Ridder et al., 2012; Hofmann et al., 2012). The present findings did not directly measure the presence of temptations or desires, but they indicate that across different social conditions, individuals high on TSC maintain a sense of having more self-control than individuals low on TSC. This could reflect their more efficient handling of conflicts overall, but it also implies that they do not calibrate their effort allocation to address conditions that require stronger responsiveness (Imhoff et al., 2014).

Agreeableness shares a modest correlation with IM (Graziano and Tobin, 2002), as was also found in the present data. However, the two constructs differ in the basic motives that account for the behavioral expressions associated with them. For the most part, Agreeableness is concerned with an emphatic concern for others’ welfare, whereas IM is focused on personal adjustment in social contexts. Notwithstanding, self-control is an important component in the development of agreeableness (Rothbart and Ahadi, 1994), and the present findings corroborate it at the trait and episodic levels. And yet, unlike IM, agreeableness was not differentially associated with levels of self-control alone and with other people. This fits our suggestion that greater self-control among individuals high on agreeableness is reserved for specific social settings, such as those involving emphatic behavior (Graziano et al., 2007).

The present study has several limitations. First, self-control was measured by self-reports and not by behavioral measures. This lends subjective value to the ratings, and thus limits to some extent their validity as in any study with self-reports. Notwithstanding, the design of the present study reduced its undesired impact because participants reported multiple times (> 27) while being in the comfort of their natural environment under terms of anonymity (vs. being observed in laboratory settings). Participants were therefore under little pressure to report in a socially desirable manner (Paulhus, 1984). Furthermore, the consistency of the present findings relating to IM with those found using alternative methods, including behavioral expressions of self-control (Uziel and Baumeister, 2012) is also reassuring that a systematic bias has little role in the association of IM with episodic self-control. An additional potential limitation concerns the generalizability of the sample. The present sample was comprised of educated young adults (mostly women) from a Western country. The theoretical claims about IM would benefit from expanding the sample composition to different populations varying in age and cultural values. Another limiting factor is the lack of specificity in the ‘with others’ context. Although this is a common approach in experience sampling research on self-control (Hofmann et al., 2012), it would have been informative to compare different compositions of social settings, especially by differentiating between settings involving greater and lesser self-presentational demands.

In summary, the present research tested a substantive theory-driven meaning to individual differences in IM under natural conditions and differentiated IM from theoretically related constructs. Its conclusions are consistent with findings from controlled experiments in showing that IM predisposes individuals to respond to public social contexts with an increase in self-control, which may lead to an overall better adjustment. These findings carry practical implications concerning the use of IM scales as validity scales in assessment processes. IM may carry (substantive) adaptive rather than maladaptive qualities and removing variance associated with IM (or participants scoring high on IM scales) from analyses (or screening processes) is likely to reduce the quality of the measurement. The present findings are in line with and extend recent approaches (e.g., Ones et al., 1996; Uziel, 2010b, 2014; MacCann et al., 2012; De Vries et al., 2014; Connelly and Chang, 2016; Müller and Moshagen, 2019; Röhner et al., 2023) that call to reconsider the application of IM scales for these purposes and shift the focus to the substantive trait being measured.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: The Open Science Framework (OSF) at https://osf.io/kqy5t/.

This study involving humans was approved by Ethics Committee of the Department of Psychology, Bar-ilan University. The study was conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

LU and TSB: Study conceptualization, data collection, data analysis, report writing. TSB: data preparation. All authors contributed to the article and approved the submitted version.

Preparation of this manuscript was supported by a grant from the Israel Science Foundation to LU (ISF grant No. 133/23).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.