The Role of Identity in Human Behavior Research: A Systematic Scoping Review

ABSTRACT

Many theories of motivation recognize a person’s identity adds value to the prediction of behavior; albeit indirectly, through constructs such as social norms, self-efficacy, perceived behavioral control, and values. Recently, behavioral research has more directly assessed the influence of identity on behavior; however, it is unclear whether such research aligns with the theoretical establishments of identity. This review investigated the identity theory alignment of existing behavior research by systematically searching, screening, and reviewing identity-behavior relevant papers, across three behavioral psychology databases. Twenty-three papers meet the inclusion criteria for review (original research with a primary focus on identity and behavior, within healthy adult population), and data relating to identity-behavior theoretical viewpoints, research methods, and study findings were extracted and synthesized to provide a scoping overview of current research. Most reviewed papers presented identity as a loosely defined construct; commonly operationalized as social identity, omitting the theorized multiplicity of social and self-identities. Overall outcomes suggested that identity influences behavior indirectly through intention strength, rather than directly as a post-intention mechanism. While existing research has been crucial in illuminating potential influences of identity on behavior, there is a need to strengthen current understandings and theoretical alignments within future research relating to identity and behavior.

KEYWORDS:

• Identity theory
• behaviour change
• self-efficacy
• integrated motivation
• subjective norms

Every day, we exist in a space that is occupied by both ourselves and others; human existence can be seen as a complex array of individual and social interactions. These interactions culminate to inform our values, the personal or social meanings we derive from various experiences, the roles we play, and who we are within a given group or situational context – that is, these interactions form our identity (Burke & Stets, 2009; Burke, 1980; Stryker, 1968). Identity theories depict identity as how we perceive ourselves in a social context (social identity; Tajfel & Turner, 1979; Tajfel, 1981; Turner et al., 1994) and as an individual person (self-identity; Berzonsky, 1994; Burke & Stets, 1999; Burke, 1980), and how these social and self-components interact together (identity theory; Burke & Stets, 2009; McCall & Simmons, 1966). Taken together, identity theories provide lenses for understanding how our views about who we are define what it means to be human, within a world where the individual exists among other individuals.

Identity theories generally agree that the identity a person comes to hold through their unique values and interpretations of the events in their life can help explain people’s behavior (Meltzer et al., 2020). That is to say, identity theories propose that people are more likely to behave in ways that hold deeper subjective meaning and value, and closely relate to their unique identity (Brandl & Bullinger, 2016). Additionally, identity theories postulate that people tend to collaboratively behave alongside other people or within their social roles and environments to accomplish a valued goal or outcome (Couch et al., 1986). Identity theories describe how people are also likely to experience cognitive distress when their identity is not externally validated and behave in ways that confirm their identity (Strachan et al., 2018). For example, if a person values learning and education but is told they are not a good student, identity theories would argue that the person may verbally argue to validate their study plan or study for longer periods of time to protect their student identity and prove their dedication to learning.

Behavioral science theories also generally agree that behavior is more likely to occur if the behavior aligns with either our own values or expectations within a social context (Bandura, 2004; Hagger et al., 2002; Webb & Sheeran, 2006). Deciding (or intending) to engage in a behavior requires a person to gather knowledge about the behavior and evaluate if the behavior is valuable to their identity, either personally or socially (Gardner & Lally, 2018; Gardner & Rebar, 2019; Rhodes & Rebar, 2017). If the behavior is deemed valuable, the person is more likely to enact the behavior and consistently maintain the behavior over time (Rebar et al., 2019; Danner et al., 2008; Gardner, 2015).

Throughout its history, numerous established behavioral science theories have also hinted at the importance of identity in human behavior. For example, the theory of planned behavior (TPB; Ajzen, 1991), suggests that people are motivated to act in accordance with their beliefs about what is socially acceptable or desirable (i.e., subjective norms), as well as what they believe their self to be capable of (i.e., perceived behavioral control). Relatedly, Bandura’s (1977) self-efficacy theory of motivation suggests a person’s behavior is driven by social interactions (i.e., vicarious experience and coaching) as well as intrinsic psychological feedback (i.e., emotional status and interpretation of experience). Another example is that of Self-Determination Theory (SDT; Deci & Ryan, 1980), which highlights a person’s need to behave in ways that meet social expectations (integrated social assimilation) and personal values (identification of self-value). More recently, the identity-value model (IVM; Berkman et al., 2017) has further highlighted identity as a crucial consideration, among the complex and various determinants of goal-directed behavior.

Both behavior and identity fields suggest an acknowledgment of the potential influence of identity on behavior. This acknowledgment fully justifies the rise in empirical identity–behavior research seeking to understand the identity–behavior relationship (see Rhodes et al., 2016; Udall et al., 2021; Vesely et al., 2021). However, amidst behavioral science, there seems to be a disconnect between the conceptualizations of identity within behavioral theories compared to the conceptualization of identity among identity theories.

The aims of the present systematic scoping review

Identity-relevant concepts are becoming more acknowledged in behavior theories and research, and a parallel line of identity theory also recognizes the role of identity in behavior. As such, it is crucial that these parallel fields recognize and align with each other’s conceptualizations of theoretical and operational bases. However, at this point, it is unclear whether, and to what degree, behavioral research reflects theoretical premises from identity theories.

The aim of our review is to move the identity-behavior field forward with clarity, by examining how behavioral science is viewing and testing identity constructs and establishing if these applications align with theoretical conceptualizations held by identity theories. Specifically, this systematic scoping review aims to investigate how identity is theoretically conceptualized, researched and measured within a broad sample of behavioral research. We will consider (1) how the field has conceptualized and measured the construct of identity, (2) how the impact of identity on behavior has been investigated, and (3) the empirical evidence regarding the role of identity in behavioral contexts. The findings of this review will inform efforts to build upon the application of the concept of identity within behavioral research. Additionally, the findings will allow for a clearer view of the present gaps in knowledge of identity and behavior, and also highlight research directions that may assist in gaining a deeper understanding of if and how identity influences behavior.

Methods

This systematic scoping review (registered with Open Science Framework Registrations; Alfrey et al., 2020) utilized the PRISMA framework (Moher et al., 2009) to consider peer-reviewed behavior research that included identity as a primary variable of interest. Specifically, the review determined the conceptualization of identity and relevant theory (if applicable) used in a broad sample of behavioral research, as well as the operationalization of the identity construct, the behavioral outcomes studied, and the key identity-behavior outcomes of the research. As there was no primary data being collected directly from humans, this study did not require Human Research Ethics application or approval.

Systematic search and screen

From July through September 2021, systematic searches were conducted in three databases known to commonly include journals relevant to behavioral psychology: Psychology and Behavioral Sciences Collection, APA PsycInfo and APA PsycArticles. The original search terms considered identity OR social-identity OR self-identity AND behavior OR behavior change OR behavior maintenance; however, this search strategy was not viable due to the inadequate and unfocused return of results. Search terms were then revised to identity AND behavio*, with expanders of equivalent subjects and limiters of peer-reviewed, full text, and document type (article). The source type was also limited to academic journals, the English language, and to the subjects of behavior and identity. The full search strategy can be found in Supplemental Materials 1. The primary search was conducted by the lead researcher (KLA); however, subsequent cross-checks were performed by additional searches (conducted by KW & JM) to validate the initial search. A total of 641 papers were produced as potential inclusions, based on the database search (see Figure 1). Of these 641 papers, 27 were removed due to being duplicates (n = 22) or inconsistently aligned with the search requirements (n = 5), leaving a total of 614 papers to be screened for inclusion.

Figure 1. PRISMA flow chart outlining systematic search, screen and review process. Adapted from Page et al. (2021).

As per best practice for systematic reviews (Moher et al., 2009), three reviewers (KLA, KW & JM) independently screened the resultant papers in two phases. The first phase was based on article title and required the title to be suggestive of 1) an original (applied) research paper with a focus on 2) identity, and 3) behavior, and 4) healthy (non-clinical) study participants over the age of 18 years, so that identity outcomes were less likely to be confounded by clinical ailment or predominant developmental processes (as seen in youth). As articles are not always able to overtly detail such criteria within their titles, each reviewer agreed to approach this first screening phase with leniency (i.e., if there was uncertainty about whether the title aligned with the criteria, the independent reviewer erred on the side of caution and accepted the title as eligible). Following independent screening, titles that were considered to meet inclusion criteria by at least two of the three reviewers were eligible for the subsequent phase of screening (majority rule). This first screening phase resulted in the exclusion of 526 papers (titles did not reasonably align with inclusion criteria pertaining to measure of identity and behavior, or non-clinical healthy adult samples; three papers were also identified as “full text unavailable” during this process) and inclusion of 88 papers into the second phase of screening.

The second phase of eligibility screening was based on the abstracts and full text of each paper (n = 88). Again, the three reviewers screened the 88 papers independently, and used the same eligibility criteria as phase one. During this second “full-text” phase, reviewers were able to understand the content of each paper more precisely, and this allowed for more precise eligibility criteria: 1) an original (applied) quantitative research paper presenting a measure of 2) identity, and 3) behavior, with 4) healthy (non-clinical) study participants over the age of 18 years. The criteria of “quantitative methods” was included to align with the reviews aim to investigate the measurement of identity within the context of behavior (as opposed to exploration of identity and behavior by way of quantitative method). Again, each reviewer screened the eligible papers independently, and papers were deemed eligible for inclusion in the data extraction phase if two of the three reviewers agreed the paper met the phase two criteria (majority rules). A total of 65 papers were excluded during this second screen (abstract and full-text review confirmed papers did not meet inclusion criteria pertaining to measure of identity and behavior or non-clinical healthy adult samples) and 23 papers were included for data extraction.

Data extraction and synthesis

The citation of each article, the theoretical conceptualization of identity, the theory informing the identity concept and the study aims were extracted, as well as the study design, measures of identity and behavior, participant characteristics and sample size, key findings and key limitations.

To ensure validity, the data extraction was conducted independently by two reviewers (KLA & KW), with any discrepancies discussed between each other until a consensus was reached. A standard Excel sheet (Microsoft Corporation, 2022) was used to collect, qualitatively analyze (e.g., thematic coded analysis) and synthesize the extracted data in its raw form. All resulting data are presented in the review results below, and within the supplementary materials files publicly available via https://osf.io/d7hjz. References for each of the 23 reviewed papers, along with the basic characteristics of the reviewed research, are summarized in Table 1 and feature an asterisk (*) within the reference list.

Table 1. Citations and basic characteristics of the 23 reviewed papers.

Citation	# Studies Presented	Identity Conceptualisation	Participant Cohort	Identity of Interest	Behaviour of Interest
Berry et al. (2014)	2	Role/social identity	University students & campus exercise classes (exercisers)	Exercise identity	Exercise behavior
Blair Evans et al. (2019)	2	Identity theory	Formal & informal running group members	Running identity	Running group engagement & running behavior
Bock et al. (2021)	1	Moral/self-identity	Emerging adults (university students)	Moral identity	Volunteer behavior
Buhrmester et al. (2015)	4	Identity theory	General Public (people with a personal connection to the Boston Bombing)	Identity fusion	Victim support actions
Chorba et al. (2012)	1	Developmental identity status	Educational psychology university students	Academic identity	Self-handicapping behavior
Fielding et al. (2008)	2	Social identity	Horticulturists & graziers	Group identification & subjective norms	Riparian zone management
Finkelstein (2008)	1	Role/social identity	Hospice volunteers	Volunteer role identity	Volunteer behavior
Finkelstein et al. (2005)	1	Role/social identity	Hospice volunteers	Volunteer role identity	Volunteer behavior
Gardner et al. (2012)	1	Identity theory	Full-time university students	Binge drinker self-identity	Binge-drinking behavior & habit
Hagger and Chatzisarantis (2006)	2	Self-idenitty	Undergraduate & postgraduate university students	Self-identity	Variety of 30 behaviors
Hardy et al. (2013)	1	Developmental identity formation & moral identity	University students	Identity formation &moral identity	Hazardous alcohol use & sexual risk taking
Hwang (2010)	1	Social identity	Undergraduate university business students	Social & self-identity	Sharing knowledge by email
Katz-Wise et al. (2010)	1	Social identity, gender role & identity saliance	First & second time parents	Work and family identity salience	Division of household labor
Kerpelman and Lamke (1997)	1	Identity microprocesses	Female undergraduate university students & their male dating partners	Career & role identity	Self-verification behavior
Marta et al. (2014)	1	Role/social identity	Young-adult volunteers working with minors	Role identity	Volunteering behavior and habit
Matsuba et al. (2012)	1	Developmental identity maturity	Environmental activists & comparison individuals	Identity maturity & Environmental Identity	Social engagement & environmental behavior
Maxwell and Visek (2009)	1	Role/social identity	Male rugby union players	Athletic identity	Competitive aggressiveness & anger behavior
Murnieks et al. (2014)	1	Identity theory	Entrepreneurs from a large metropolitan area	Identity centrality & salience	Entrepreneurial behavior
Nigbur et al. (2010)	2	Social identity	Households within Borough Council of Guildford	Neighbourhood identification & self-identity	Recycling behavior
Smith et al. (2008)	2	Self-identity	Introductory psychology university students	Self-identity	Consumer behavior
Tyler (2009)	1	Social identity	General population	Superordinate & ethnic identification	Law abiding behavior
White et al. (2008)	1	Role identity/identity theory	Undergraduate psychology university students	Student role identity	Students’ peer-assisted study session attendance
Wilson and Muon (2008)	1	Social & role identity	Physical education & kinesiology university students	Exercise identity	Exercise behavior

Results

The primary focus of this review is to investigate how identity has been considered, researched, and measured within a broad sample of human behavior research. The main results of interest focus on the theoretical conceptualization of identity, how identity and behavior has been studied (research approaches and construct measurement), and study outcomes relating to identity–behavior relationships. The following results are based on text data extracted from the 23 reviewed papers (a total of 32 studies) that align with the criteria of interest.

The conceptualization of identity

Research aims and theoretical conceptualizations of identity were extracted and analyzed per paper and are thereby reported at the paper level (i.e., 23 papers). The aims and theoretical viewpoints presented by each of the 23 reviewed papers are detailed within Supplementary Materials 2 (https://osf.io/d7hjz).

Research aims

Most of the 23 reviewed papers presented aims of investigating associations (9/23) or predictive (8/23) relationships between identity and behavior. A few reviewed papers (4/23) aimed to validate new or existing identity measures or test the unique value of including identity constructs within existing conceptual models. A few reviewed papers (2/23) aimed to integrate traditionally separate existing approaches (identity and functional analysis; passion and identity theory) to better understand behavior.

Theoretical conceptualisations of identity

All reviewed papers clearly presented identity-related theoretical conceptualizations of identity; however, some lacked clear details of how the theoretical viewpoint aligned with the research aims (e.g., the aim of the study was not overtly supported by a relevant identity theory approach). Within the 23 reviewed papers, the most common use of “identity” was aligned with social identity (11/23), suggesting social values, experiences and expectations are key factors in identity and behavior (e.g., Stryker, 1987; Taifel & Turner, 1986; Tajfel & Turner, 1979). Such socially relevant constructs were presented as “social identity,” “role identity,” “exercise identity,” and “athletic identity.” However, the alignment of social identity components and ideas was not always presented with clear detail or definition. For example, one paper presented the use of “Stryker’s identity theory” (i.e., assumedly preference toward structured social identity); however, the research stated a focus on self-identity with social norms as an additional construct. This undefined mixing of social and self-identity suggests a potential misalignment with the use of the separate (yet connected) identity theories.

Identity theory, as an overarching theory (e.g., Blumer, 1986; Burke & Stets, 2009; McCall & Simmons, 1966), was also a common conceptualization (6/23). In these papers, identity was often generically described as involving a person’s values, roles and social contexts; however, the notion of reciprocal social- and self-elements of identity – a key tenet of Identity Theory – were not always clearly presented. For example, some papers did not present a clear argument as to why the use of overarching identity theory (being the interaction of social and self) was appropriate to their research aims, or why they used the overarching identity theory rather than narrowing into (for example) social or self-identity.

Developmental identity, as presented by Erikson (1963, 1968) and Marcia (1966), is traditionally aligned with developmental theories rather than identity theories, as it describes the biopsychosocial formation and status of identity during adolescence and very early adulthood. However, developmental identity featured as the conceptualization of identity in a small number of reviewed papers (3/23) that focused on undergraduate students and youth. Within the reviewed papers using developmental identity concepts, more mature and stable identity (i.e., identity achievement and commitment to a sense of identity) was most often used to predict behavior.

Self-identity (e.g., Berzonsky, 1994; Burke & Stets, 1999; Burke, 1980) was a less commonly presented identity viewpoint (3/23). One reviewed paper presented self-identity in terms of moral identity (i.e., moral goals, values and commitments); however, other reviewed papers presenting moral identity did so in collaboration with developmental identity constructs. Other self-relevant identity approaches considered identity as self-prescribed labels (or roles) or a dichotomous self-categorization of gender, sexuality, or career type. In all, there appeared to be somewhat of a misalignment with the theoretical basis of self-identity as personal values and beliefs developed from personal experiences and subjective interpretations.

Research approaches

Study design and participant characteristics were extracted from each study presented within the 23 reviewed papers (a total of 32 studies). Thereby, analysis and subsequent results are reported per study, rather than per paper. Supplementary Materials 3 (https://osf.io/d7hjz) details the extracted study designs and participant characteristics data.

Study design

Of the 32 studies presented, the majority of studies utilized a cross-sectional (19/32) or prospective designs (12/32), few presented a mixed methods approach (4/32), and only one paper presented an experimental approach. Of the 12 prospective studies, most (10/12) comprised two time points varying from 1 week apart to 3 years apart, while one study comprised four timepoints and one was unclear about the length of time between the timepoints. All papers presented self-reported data collection, with only a few also utilizing additional methods including observational (2/32) or coded qualitative (1/32) data, alongside the self-report data.

Participant characteristics

Participant cohort characteristics remained reasonably consistent within the papers, even where multiple studies were presented. Participant samples were predominantly university student populations (14/32) or behavior-specific populations (11/32) such as exercisers, running groups, farmers, volunteers, parents, environmental activists, rugby union players, and entrepreneurs. Few general population samples were used and were recruited via participant databases (5/32) or by specific local government area data (3/32). In alignment with the systematic inclusion criteria, all reviewed studies focused on healthy adult participants (aged 18+) who were not noted to be of any clinical status.

Measurement of identity and behaviour

Amongst the 32 studies, identity and behavior measures were either a direct measure (one scale directly measuring construct of interest) or a battery of construct-related scales (multiple scales measuring related constructs to infer level for construct of interest). Behavioral measures also included hypothetical (“would you usually do this” or vignettes) or intentioned (“in the next week, do you intend to do this”) measures of behavior. Due to variation in the number and types of measures across the 32 studies, the following results consider whether measures were validated, adapted or created, and the focus or type of construct measured (type of identity; type of behavior) and are presented in terms of frequency. An overview of the identity and behavioral measures presented across the 32 studies is detailed in Supplementary Materials 4 (https://osf.io/d7hjz).

Measurement of identity

Identity constructs were mostly measured with either validated scales or subscales (f = 15), or adaptions of validated scales (f = 12); for example, the Exercise Identity Scale (Anderson & Cychosz, 1994) and the adapted runner identity scale (adapted from Strachan & Brawley, 2008 and the Exercise Identity Scale from Anderson & Cychosz, 1994). In addition, some measures (f = 5) were specifically developed to suit the context of the study (e.g., two items were developed to specifically assess soft-drink consumer identity), and some (f = 2) of those developed scales were based on dichotomous or categorical items to categorize identity (e.g., superordinate and ethnic identification). In addition, some studies also recruited participants specifically from the behavioral setting in which they were likely to hold an established identity for, for example, a study of running identity within running groups and exercise settings. Many studies displayed a predominant focus on socially driven identity constructs such as role (f = 9) and social identity (f = 6), whereas few placed predominant focus on personal identity constructs, such as self-identity (f = 3), developmental identity (f = 3), and moral identity (f = 2).

Measurement of behaviour

Many studies used measures specifically developed to assess frequency and/or occurrence of the behavior of interest (f = 27), and three of those developed measures were dichotomous categorization of behavioral engagement (e.g., law-abiding behavior). However, frequency and/or occurrence of behavior was also frequently (f = 13) measured with validated scales or adaptions of validated scales (e.g., Exercise Godin Leisure-Time Exercise Questionnaire from Godin & Shephard, 1985). There were also instances of indirect measures of behavior through “usual behavior” and vignettes or behavior intentions (f = 7). Most behavioral engagement measures were self-reported reflections of past behavior; however, there were also some instances of researcher-observed behavior and behavioral coding (f = 3).

The most evident behaviors of interest included social behaviors (f = 7; e.g., social support giving, sharing of knowledge, division of household labor, social engagement, law-abiding, social aggression, and peer support session attendance), volunteer behavior (f = 4), physical activity or exercise behavior (f = 3), health behavior (f = 3; e.g., alcoholic drinking and sexual risk-taking), pro-environmental behavior (f = 2; farming and recycling), self-verifying behavior (f = 2; self-alignment and self-handicapping), consumer behavior (f = 1), and entrepreneurial behaviors (f = 1).

Research outcomes

Research outcomes (key findings and limitations) were extracted and analyzed per paper and are thereby reported at the paper level (i.e., 23 papers). An overview of the findings related to identity and behavior is presented in Supplementary Materials 5 (https://osf.io/d7hjz).

Key findings

All but one of the reviewed papers presented findings of identity being positively related with behavior (22/23). These positive effects were predominantly suggested via the measured effect of identity on intention formation, which, in turn, was assumed to be the predictor for behavior (11/23). Mature, stable, and strong identity was particularly related to positive behavioral intention and behavior (3/23). Social identity factors were commonly seen to predict engagement in socially interpreted behaviors (such as running group engagement and recycling behaviors; 6/23). Only one study presented a null effect of identity on behavior, with results showing no evidence of athletic identity being a significant predictor of aggressive behavior. This may be an exception, however, as athletic identity has not always been described as aligned with mature or stable identity (Brewer et al., 1993), and the majority of the positive relationships found in this review were based around the presence of mature and stable identity.

Research limitations

The reviewed papers often noted the limited generalization of study findings due to small or restricted participant samples and limited ecological insight (8/23), as well as the potential for biases relating to self-reported data (7/23). Participant samples were also frequently convenience samples comprising university students (11/23) or cohorts that identified with specific behaviors (8/23; e.g., running identity, exercise identity, volunteer identity). Additional limitations not mentioned by the papers, but noted during analysis, included that causal inference could not be assumed due to the cross-sectional nature of some studies, loose theoretical definitions, intentional or hypothetical behavior (rather than enacted behavior), and lack of measurement in specific behavioral stages.

Discussion

The aim of this review was to examine how behavioral science was viewing and testing identity constructs and establish if these applications aligned with theoretical conceptualizations held by identity theories. A sampling of the theoretical conceptualizations of identity, research approaches and key outcomes of identity-behavior research is reviewed. The use of identity theory was apparent in all reviewed papers but, at times, was presented as a single, loosely described concept, which was sometimes misaligned with the operationalization of identity measured within the research. The primary aim of reviewed research was to investigate potential relationships between identity and behavior directly. Study designs were largely cross-sectional, and data tended to be from convenience sampled populations who may have held existing salient identities or knowledge about the behavior of interest. The measurement of both identity and behavior was most often collected via self-reported survey. Overall, the studies reported findings of a positive relationship between identity and behavior. Some findings showed predictive links between social identity and social behaviors, suggesting there may be differing roles for differing types of identity within certain behavior settings.

Theoretical conceptualisations of identity

The aims of the reviewed studies were predominantly focused on direct associations or predictions of identity and behavior (including behavioral intention and hypothesized behavior). Validation of new and existing measures and models, and blending of existing approaches, were also used to understand the role of identity in behavior contexts. Such approaches suggest that behavioral research is in the early stages of investigating the value of identity as a potential behavior determinant, at least as is depicted by the sampling of research in this scoping review. The loose conceptualization and definition of identity within the reviewed papers also suggests that identity-behavior research is in the early stages of understanding the complexities of the various identity theories and how they fit within existing behavioral science theories and models. For example, while some papers acknowledged various theoretical conceptualizations of social and self-identity (e.g., moral identity, role identity, exercise identity), others relied heavily on generic description, assumed knowledge, and/or citations of previous publications to explain their use of identity. The identity-behavior field should aim to first deepen our understandings of identity theory in terms of behavior and then investigate identity-behavior relationships to move this research field forward and produce more developed and rigorous identity-behavior evidence.

Many reviewed papers conceptualized identity with a lens of social identity (Tajfel, 1981; Tajfel & Turner, 1979; Turner et al., 1994). This social focus is a common occurrence in current identity-behavior research (e.g., Brenner & DeLamater, 2013; Davis et al., 2019; Pan et al., 2019) as it aligns well with behavioral theories such as Theory of Planned Behaviour (TPB; Ajzen, 1991) and Social Cognitive Theory (SCT; Bandura, 2001). However, as identity comprises a complex reciprocal relationship between social and self-view, it is reasonable to suggest that identity-behavior research should also acknowledge aspects of self-identity (Berzonsky, 1994; Burke & Stets, 1999; Burke, 1980).

Identity Theory (Burke & Stets, 2009; McCall & Simmons, 1966) is commonly accepted as an overarching view of this reciprocal relationship between social and self-views, and some papers in this review did present “identity theory” as their theoretical viewpoint; however, there was very little acknowledgment of self-identity aspects, or the reciprocal nature, that relates to this overarching identity theory. It is evident that, even when using theory that suggests reciprocal social and self factors, much of the focus in current identity-behavior research remains on the social aspects of identity.

Some of the reviewed papers did venture outside of the typical social identity viewpoint; however, many of those papers did not delve into the theoretical underpinnings of the chosen theory or viewpoint. Of the few reviewed papers that acknowledged self-identity, most contextualized within the domain of “moral identity” without clearly explaining the theoretical and multifaceted nuances of moral or self-identity. Similarly, in some reviewed papers, dichotomous self-identification items were adopted to allow identity categorization; this approach results in relevant “social grouping” categorizations, but seemingly neglects the salience continuum and reflexive nature of identity (Stets & Burke, 2000; Stets et al., 2020).

The inclusion of developmental identity (Erikson, 1963, 1968; Marcia, 1966) was also noteworthy, as developmental identity is more often associated with identity-formation factors and determinants, rather than how identity may influence behavior. However, the use of developmental identity formation in the reviewed papers was clearly linked to the research aims and methods, as these studies focused on identity formation in younger cohorts, and how the maturity status of identity influenced behavior. Indeed, developmental identity in younger adulthood may be important to consider as the investigation of identity–behavior relationships continues to evolve, as identity formation during younger years is shown to hold flow-on consequences for middle and later adulthood (Erikson, 1963, 1968; Marcia, 1966). In all, despite all identity theories generally acknowledging various levels of social- and self-aspects of identity (Burke & Stets, 2009), the role of identity in behavior research, as represented by this scoping review, has focused heavily on the social aspects of identity and is limited in its understanding of identity viewpoints beyond the social experience.

Methodology of identity-behaviour research

Throughout the reviewed papers, cross-sectional study designs were most used, meaning that resultant outcomes were mostly limited to correlational insights. Brief prospective designs were also commonly used; however, most of these were limited by only two timepoints and/or with differing measures at each timepoint (e.g., identity measured at timepoint 1 and behavior measured at timepoint 2), which may have elicited a priming effect for engagement in the measured behavior. Overall, the research approaches employed by the reviewed papers were able to present key “snapshot” insights addressing the potential role of identity in behavior contexts but lacked insight into causality of these effects or consideration of dynamics of identity or behavior beyond change over two time points.

In terms of data collection, most variables were self-reported via previously validated or adapted measures, and based on reflective cognition (e.g., do I value this? or how long did I do this behavior?), intended behavior, or hypothetical behavior within vignettes. That identity is linked to behavioral intentions and is aligned with evidence such as a recent review by Vesely et al. (2021) which found strong links between pro-environmental intentions, behaviors, self-identity, and social identity. Measures based on intention and hypothetical behavior may not be adequately reflective of actual behavior, as intention is not always a stable predictor of behavior (Rebar et al., 2019; Dombrowski et al., 2014; Rhodes & de Bruijn, 2013). A few of the reviewed studies reduced the risk of response bias by employing data collection via field observations and qualitative audio recordings (which were subsequently coded for quantitative analysis purposes); however, most were self-reported measures. Future identity-behavior research should aim to directly assess behavioral outcomes, rather than degrees of future planned action or retrospective self-reported behavior.

A few of the reviewed papers also assessed identity via self-selected group categorization or membership (e.g., gender and ethnic identity). Self-selected group membership is a simple means to identify whether someone relates themselves to a given context, and such memberships may infer common in-group/out-group connotations (e.g., gender X may hold different connotations compared to gender Y). However, without exploration of such connotations, the dichotomous categorization approach overlooks the continuum and complexity of variables that underpin identity. Also, as many of the reviewed studies were based on samples directly relating to the identity and behavior of interest (e.g., exercisers, students, runners, rugby players, etc,), it should be considered that those participants are likely to hold preexisting salient identity for that relevant behavior (and therefore self-identified as being eligible for the study), potentially resulting in an underrepresentation of the lower spectrum of the identity salience continuum. In their review, Udall et al. (2021) also found environmental (lab vs field) and cohort (student vs non-student) differences in pro-environmental identity research, which potentially support the data collection and cohort effect concerns raised within this current review. To fully understand identity and its influence on behavior, samples need to comprise individuals with a broad array of identity levels and behavioral engagement.

The methodological designs of the reviewed papers were also commonly limited by the cross-sectional or brief correlational prospective approaches used in the research. Such “snapshots” can provide foundational understandings pertaining to between-person relationships, but they are limited in their capability to track ecological factors or fluctuations in variables across time and setting. The variables, methods and outcomes presented within the reviewed papers are valuable in understanding core identity–behavior relationships; however, to progress this understanding further, more methodologically complex approaches (e.g., ecological momentary assessment designs; Rebar et al., 2020; Shiffman et al., 2008) may be needed to narrow in on the more precise interactions between identity and behavior.

Limitations of the current review

It is acknowledged that this review only considers a small sample of identity-behavior related research. However, the aim of this review was to provide a synthesis of a sampling of how identity has been considered within behavior research, to consider the alignment of identity theory within behavior research approaches. The sample of 23 reviewed papers resulted organically from conservative yet systematically structured search and screen processes, which can be considered a valid approach for such a holistic yet broad synthesis. The decision to not focus on any particular behavior or identity type was also purposive, to support this aim of a holistic overview of identity-behavior research to date. Just as valuable previous reviews have done (Udall et al., 2020, 2021; Vesely et al., 2021), future reviews may consider narrowing to a distinct behavior or identity type, to directly synthesize the role of various identity types and/or various types of behavior or stages of behavior processes (e.g., intention, engagement, and maintenance).

Limitations regarding the search and screen criteria of full-text access to articles should also be acknowledged. The full-text accessible criterion was considered a necessary approach for full-text screening and data extraction processes. However, it should be noted that full-text access may differ from institution to institution, and this may limit the replicability of this review. This particular limitation is a relevant concern for all fields of research and echoes the call for accessible options regarding open access publication for aggregate works such as reviews.

There was also some potential for misinterpretation or researcher biases during screening and data extraction processes. To minimize this risk, three independent reviewers were involved in the screening and eligibility phases of the review, with majority rules being employed in case of inclusion disagreement, and data extraction was completed by two independent reviewers to ensure coverage of data. The PRISMA framework (Moher et al., 2009) was also used to structure the review in a valid and empirical method.

Finally, this systematic scoping review did not consider qualitative findings or other reviews of the identity topic, due to the focus on investigating how applied behavior research has measured (rather than explored) the concept of identity. Future reviews may look to consolidate qualitative insights for a richer overview of perceptions pertaining to the identity-behavior relationship or focus on meta-analyses to boost quantitative generalizability.

Conclusions

Overall, this review highlights a broad sample of research that illuminates the multifaceted and valuable role of identity in behavior. The outcomes of this review align with various other reviews that explore the role of identity and identity-related concepts in specific behavioral domains (e.g., Rhodes et al., 2016; Udall et al., 2020). However, this review extends to sample the empirical conceptualizations and research methods used within identity-behavior investigations and highlights the need for improved conceptual clarity and theoretical alignment of both identity and behavior. This review also aligns with the suggestions of identity-behavior complexities and the power of identity’s role in behavior, as presented by Berkman et al. (2017).

The identity-behavior research conducted to date has revealed a valuable relationship between identity and behavior; however, further research acknowledging various theoretical and conceptual complexities of identity and behavior is required to better define what that relationship looks like. While continuing to investigate identity–behavior relationships and implications, it is essential that conceptual and theoretical alignment is maintained, to ensure findings can be interpreted with conceptual, theoretical, and practical validity.

Calls for future research agenda

This systematic scoping review highlights a relationship between identity and behavior, within behavioral research contexts. However, the conceptualization and operationalization of identity for behavioral research fields remains narrow, given that the focus is predominantly on social aspects of identity and tends to be tested as direct associations rather than as theorized multivariate processes. Future identity-behavior research should consider stepping beyond the social identity focus to acknowledge the diverse theoretical bases of identity and explore the potential influences of self-identities on behavioral processes and to test these associations amidst relevant theories. Acknowledging and integrating these theoretical establishments in future identity-behavior research will not only deepen our understanding of multifaceted identity within behavioral contexts, but it may also clarify specific identity determinants (e.g., social vs self) within certain stages of behavioral processes (e.g., intention vs maintenance).

Exploring individual interpretations of identity and how it features in real-world contexts may also uncover if identity is perceived as a major contributor to the decision to engage in a behavior and may also clarify if identity is experienced as a social and reflective process (as suggested by Reynolds et al., 2020; Tarrant et al., 2020) or a non-conscious automatic process (as suggested by Rhodes, 2017), or as a reflexive process that automatically shifts over time and context (Michie et al., 2011; Rhodes et al., 2021, 2023). Consideration of varying levels of identity and identity salience should also be a focus for future identity-behavior research, as individuals with low or no preexisting behavioral identity may report different experiences to individuals who already firmly identify with the behavior of interest.

In developing a more informed understanding of identity relating to behavior, we can better understand the personal and social meanings and values within behavioral contexts, and bring the identity-behavior research field forward whilst ensuring alignment with the theoretical establishments of identity. Mechanisms of identity should highlight the origins of identity and the theoretical structure that underpins identity (e.g., varying identities and salience) but also consider how identity is experienced (e.g., reflective vs automatic), and how it impacts on behavior in the real-world (e.g., interventions to encourage positive behavior change via identity shifts). Such a multifaceted and sound understanding of the role of identity within behavioral processes may provide alternative, or complementary, approaches to encourage positive behavior for people striving to become the best possible versions of themselves.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the Australian Government Research Training Program Stipend Scholarship CQ University Australia Research Higher Degree Candidate Funds .