In the last century, many psychologists tried to identify different facets of an individual’s self-concept, such as two actual selves from the perspective of individuals themselves and others (Erikson, 1959; Wylie, 1979), spiritual self and social self (Pedersen, 2000), possible selves and now selves (Markus and Nurius, 1986; Leung, 2011), and so on. It is a long-accepted notion that an individual’s self-concept consists of different self-state representations. Higgins (1987) integrated different facets of self-concept proposed by previous theorists and proposed self-discrepancy theory, which provides a systematic and comprehensive framework to reveal the interrelations among different self-states.

Self-discrepancy theory (Higgins, 1987, 1989) postulates three basic domains of the self (i.e., actual self, ought self, and ideal self) and two basic standpoints on the self (people’s own personal standpoint and the standpoint of some significant others). More specifically, the actual self represents the attributes that individuals or their significant others believe they actually are. The ideal self and the ought self represent the attributes that individuals or their significant others think they can ideally possess and ought to possess, respectively. In addition, the combination of the ideal self and the ought self was defined as an individual’s self-guide, which is the standard for self-evaluation and self-enhancement in daily life (Higgins, 1987). Self-discrepancy refers to the gap between the actual self and the self-guide. According to self-discrepancy theory, the greater an individual’s self-discrepancy is, the more discomforts he or she will suffer (Higgins, 1987, 1989). It is suggested that people’s behavior is oriented to reduce self-discrepancy (Hu et al., 2015), thereby relieving discomfort and negative emotions.

Individuals may present themselves differently in different contexts (Gee, 2000; Marwick, 2013). The phenomenon that people selectively present themselves in different situations was defined as “strategical self-presentation” (Ranzini and Lutz, 2017; Jang et al., 2018). With the behavior of strategical self-presentation, people can reduce their self-discrepancy to some extent. For instance, an unattractive person can present a better self-image and reduce actual-ideal self-discrepancy by wearing fashionable clothes.

The development of various online platforms (such as social network sites and online games) allows people to present themselves strategically to a greater extent. Individuals can construct their identity in virtual spaces and create a virtual self-online freely (Suh, 2013). An individual’s virtual self is likely to be different from the actual self. For example, on social network sites, people may manipulate their photos to build a more desirable image. In online games, players can pick or construct different game avatars (Loewen et al., 2021). It is found that players perceive their game avatars as more extroverted and less neurotic (Bessière et al., 2007). People also rate the virtual avatar in Second Life as more outgoing and attractive than themselves (Messinger, 2007). Scholars proposed the concept of virtual self-discrepancy to describe the degree to which an individual’s virtual identity deviates from the individual’s actual identity in the real world (Jin, 2012). Individuals may experience virtual self-discrepancy when they engage in online activities. Previous studies suggested that virtual self-presentation has a significant influence on people’s emotions and behavior. Thus, the current study will review the effect of both traditional self-discrepancy and virtual self-discrepancy on online behavior.

To identify studies relevant to the effect of self-discrepancy on online behavior, five major databases were used in the literature search phase, including Scopus, ScienceDirect, Web of Science, Academic Search Complete, and Wiley Online Library. These databases cover a wide range of disciplines and a large number of peer-reviewed journals. In the search process, the terms related to self-discrepancy (such as “self-discrepanc*,” “actual-ideal,” and “actual-ought”) were combined with terms related to the online context (such as “online,” “cyber,” “Internet,” and “virtual”) as keywords. Full-text searches were conducted to cover as many studies as possible. The search was conducted in December 2021. There was no limitation on the time interval of the publications. We searched for possibly relevant articles published by the date of the literature search, yielding a total of 2,570 papers.

To select relevant articles from the search result, a set of eligibility criteria were established. Studies were included if they meet all the following inclusion criteria: (a) investigate the effect of self-discrepancy; (b) involve an online context; (c) published in peer-reviewed journals; and (d) written in English. The exclusion criteria ensure that studies were not (a) conference papers, book chapters, and review papers; (b) only focusing on the offline context. Both qualitative and quantitative studies were considered in this review as long as they examined the effect of self-discrepancy.

After removing duplicates across databases, 2,157 articles were left. Next, the titles and abstracts of the studies were checked. Most studies (n = 2,017) were removed based on the inclusion and exclusion criteria. Then, the authors went through the full text of the remaining 140 articles to access their relevance based on the eligibility criteria and research questions. One hundred three studies were further excluded. Finally, 37 articles were included for this review. Figure 1 presents the flowchart of the selection process.

Table 1 presents a summary of the basic characteristic of the reviewed articles. As shown in Table 1, a total of 13 studies (35.1%) used a student sample. Three studies (8.1%) used a mixed sample of students and non-students. The remaining 21 studies (56.8%) used a non-student sample or did not illustrate the characteristics of their participants clearly. The sample size ranged from 33 to 1,063.

Regarding the research context, 14 articles were conducted on social network sites. Most of them investigated a specific site, such as Facebook, Instagram, and QQ. Nine studies were carried out in the context of online games. Six articles focused on self-discrepancy in online communities. Other research contexts include online shopping, social media (such as YouTube), the general online world, and so on.

Most reviewed articles used quantitative approaches. Only two studies adopted a qualitative research design. The most frequently used research methods are surveys and experiments. Experiments are often used in research conducted in the context of social network sites and online games. Scholars designed different experiments to investigate how self-discrepancy is associated with social network use and gaming behavior.

When it comes to the measurement of self-discrepancy, there are no fixed measurements. The instruments used in the reviewed articles can be classified into two categories: direct measurement and indirect measurement. Some studies assessed participants’ self-discrepancy directly. They used general items to evaluate participants’ overall perception of the discrepancy between different self-concepts (e.g., “In this online community, I am smarter and more intelligent than I am in the real world”). Some studies assessed participants’ self-discrepancy indirectly. Given that different people have different actual self and self-guide, some studies asked the participants to report different aspects of their self-concepts. Then, self-discrepancy was measured with the differences in scores. When using the indirect measurement, some research asked the participants to think about adjectives to describe their self-concept by themselves, while some other research provided a list of personality traits or description sentences based on the research themes and asked the participants to rate the items in the list from different perspectives (e.g., the extent to which the item reflects their ideal self, virtual self, or actual self).

The emergence of the Internet has changed the way people express themselves and interact with others (Huang et al., 2019), which may affect people’s self-discrepancy on the Internet. Individuals may take advantage of the characteristics of the Internet and present themselves selectively online. Additionally, individuals can access various information and communicate with different people easily in the online world. The new information and new social connections online may influence their self-guide. In this case, the Internet may affect people’s self-discrepancy in the online world by changing the way people present their self-concept and the formations of their self-guide.

The self-discrepancy theory proposes that different types of self-discrepancies will result in different kinds of psychological discomforts, including dejection-related emotions (such as disappointment, dissatisfaction, and shame) and agitation-related emotions (such as fear, guilt, and self-contempt; Higgins, 1987). An individual’s behavior is oriented to align their self-concept with self-guide to avoid those negative emotions (Higgins, 1987). When experiencing negative emotions induced by self-discrepancies, some people may engage in non-deviant behavior and try to overcome these psychological discomforts through reducing self-discrepancy. By contrast, some may engage in deviant behavior and try to escape from these discomforts temporarily by ignoring self-discrepancy.

The findings of this review should be evaluated in the context of limitations. This study only considered peer-reviewed articles written in English in the review process. Therefore, the findings of this study might not be comprehensive enough. Future reviews are suggested to include studies published in other venues (such as book chapters and conference proceedings) and articles written in other languages.

Although this review has some limitations, it contributes to the research on self-discrepancy in several ways. This study presents the basic characteristics of the included studies, such as sample characteristics, research context, and data collection method. It elaborates on the impact of the Internet on people’s self-discrepancy. It is found that people may experience a decrease in self-discrepancy when they present themselves in the way they want online. Individuals may perceive an increase in self-discrepancy when they integrate online information and values into their self-guide. In addition, this study summarizes the effects of self-discrepancy on different kinds of online behavior. Results of this review provide an empirical landscape of the research on self-discrepancy in the Internet era. Also, this review identifies some gaps in existing studies and provides several recommendations for future research.