The results of the search and screening process leading to the selection of the studies included in the review are shown in Fig. 2. The data extracted from the studies, as well as their methodological quality, are summarised in Table 1.
Of the twelve articles included in the results, ten are quantitative studies and two are qualitative. Of the ten quantitative studies, seven are correlational studies (Bernhold, 2019; Bernhold & Metzger, 2018; Liebers & Schramm, 2017; Rain & Mar, 2021; Ramasubramanian & Kornfield, 2012; Schmid & Klimmt, 2011; Tian & Hoffner, 2010), two are randomised experimental studies (Sestir & Green, 2010; Teasdale et al, 2021) and one used a correlational method with one sample and a randomised experimental method with another (Shedlosky-Shoemaker et al, 2014). Both qualitative studies consist of thematic analysis (DeGroot & Leith, 2018; Karhulahti & Välisalo, 2021).
Of the total number of studies, six studied parasocial relationships (Bernhold, 2019; Bernhold & Metzger, 2018; DeGroot & Leith, 2018; Karhulahti & Välisalo, 2021; Liebers & Schramm, 2017; Schmid & Klimmt, 2011; ), four studied identification (Ramasubramanian & Kornfield, 2012; Sestir & Green, 2010; Teasdale et al, 2021; Tian & Hoffner, 2010;) and two studied both (Rain & Mar, 2021; Shedlosky-Shoemaker et al., 2014 ).
Flow chart of the screening process
Table 1 Summary of data extracted from selected studies
References
|
Sample
|
Design
|
Object of Study
|
Instruments
|
Quality (MMAT)
|
Sestir & Green, 2010
|
N = 127
Gender: 59.84% Women; 40.16% Men
Age: Unknown
Country: USA
|
Quantitative, randomized experimental study
|
Effect of identification with fictional characters and transportation on self-concept
|
Me/Not-Me Task;
Ad-hoc Likert scale on identification and transportation;
Ad-hoc Likert scale on character traits
|
5/5
|
Tian & Hoffner, 2010
|
N = 174
Gender: 73.4% Women; 26.6% Men
Age: 18–64 (M = 30.5; SD = 11.3)
Country: USA
|
Quantitative, correlational study
|
Relationship between perceived similarity, identification and parasocial interactions with striving to resemble a fictional character
|
Parasocial Interaction Scale (Rubin & Perse, 1987);
Television Affinity Scale (Rubin & Perse, 1994);
Audience-Persona Interaction Scale (Auter & Palmgreen, 2000);
Identification Scale (Cohen, 2001); Ad-hoc Likert scale to assess change/influence
|
5/5
|
Schmid & Klimmt, 2011
|
N = 1358
Gender: 61% Women; 39% Men
Age: 10–56 (M = 20.2; SD = 7.8)
Country: 50% Germany; 50% Mexico
|
Quantitative, correlational study
|
Cross-cultural differences in parasocial relationships with fictional characters
|
Ad-hoc semantic differential scale on character perception;
Ad-hoc Likert scale on parasocial Relationships;
Admiration/Attraction Scale (Visscher & Vorderer, 1998).
|
4/5
|
References
|
Sample
|
Design
|
Object of Study
|
Instruments
|
Quality (MMAT)
|
Ramasubr
amanian &
Kornfield,
2012
|
N = 385
Gender: 71.43% Women; 24.15%
Men; 4.42% not especified
Age: +18 years old (M = 25.69;
SD = 7.54)
Country: USA
|
Quantitative,
correlational
study
|
Features of the relationship between Japanese shōjo and the American public, on the basis of cultural differences
|
Ad-hoc Likert scales on character perception, wishful identification and parasocial interactions
|
5/5
|
Shedlosky
-Shoemak
er et al,
2014
|
N = 113
Gender: 65.49% Women; 34.5%
Men
Age: 18–28 (M = 18.8;
SD = 1.35)
Country: USA
|
Quantitative,
correlational
study
|
The influence of transportation into the story, inclusion of others in self, cognitive overlap and perceived similarity to fictional characters to current and ideal self, as means for personal growth
|
Inclusion of Other in Self Scale
(Aron et al., 1992);
Transportation Scale (Green &
Brock, 2000);
Self-Expansion Questionnaire
(Lewandowski & Aron, 2002);
Ad-hoc Likert scale about Character
Perception
|
4/5
|
N = 79.
Gender: 44.3% Women; 55.7% Men
Age: 18–36 (M = 19; SD = 2.57). Country: USA
|
Quantitative, randomized experimental study
|
4/5
|
Liebers &
Schramm,
2017
|
N = 493
Gender: 79% Women; 21% Men
Age: 14–74 (M = 27.7; SD = 10.3)
Country: Germany
|
Quantitative,
correlational study
|
Relationship between type of character attractiveness and reading experience with parasocial relationships with fictional characters
|
Ad-hoc Likert scales on parasocial relationships, character attributes and reading experience, created by merging different existing scales
|
3/5
|
References
|
Sample
|
Design
|
Object of Study
|
Instruments
|
Quality (MMAT)
|
Bernhold
& Metzger,
2018
|
N = 261
Gender: 67% Women; 30%
Men, 3% not specified
Age: +55 (M = 62.3; SD = 5.27)
Country: USA
|
Quantitative,
correlational study
|
Relationship between depressive symptoms and parasocial relationships with fictional characters, anxious attachment and the quality of relationships with partners and offspring
|
Center for Epidemiologic Studies
Depression Scale (Radloff, 1977);
Marital Opinion Questionnaire
(Huston et al., 1986);
Experiences in Close Relationships-Revised Scale (Fraley et al., 2000); Virtual Friendship Scale (Hartmann et al., 2008)
|
4/5
|
DeGroot &
Leith,
2018
|
Public messages (N = 232) from
Facebook were analysed without collecting information on the users who posted them.
|
Qualitative,
thematic
analysis
|
Parasocial grief following the death of a fictional character and its expression in social networks
|
No instruments were used
|
5/5
|
Bernhold,
2019
|
N = 206
Gender: 64.6% Women; 33%
Men; 2.4% not specified
Age: 55–77 (M = 61.9;
SD = 4.42)
Country: USA
|
Quantitative,
correlational study
|
Relationship between depressive symptoms and loneliness with hostile parasocial relationships with fictional characters, the quality of romantic relationships, and anxious and avoidant attachment
|
Revised UCLA Loneliness Scale
(Russel, 1996);
Marital Opinion Questionnaire (Huston et al., 1986);
Experiences in Close Relationships-Revised Scale (Fraley et al., 2000); Antipathy scale (Hartmann et al., 2008);
Virtual Friendship Scale (Hartmann et al., 2008)
|
5/5
|
References
|
Sample
|
Design
|
Object of Study
|
Instruments
|
Quality (MMAT)
|
Karhulahti & Välisalo, 2021
|
Public messages (N = 1667) from forums were analysed without collecting information on the users who posted them.
|
Qualitative, thematic analysis
|
Parasocial romantic and sexual relationships
|
No instruments were used
|
5/5
|
Rain & Mar, 2021
|
N = 150
Gender: 56% Women; 44% Men
Age: 17–28 (M = 19.3; SD = 1.89).
Country: Canada
|
Quantitative, correlational study
|
Relationship between attachment style and personality, and identification, parasocial interactions and parasocial relationships
|
Attachment Style Questionnaire (Feeney et al., 1994);
Big Five Inventory (John & Srivastava, 1999);
Parasocial Interaction Scale (Cole & Leets, 1999);
Identification Scale (Tal-Or & Cohen, 2010);
Experience of Parasocial Interaction Scale (Hartmann & Goldhoorn, 2011)
|
5/5
|
Teasdale et al., 2021
|
N = 83
Gender: 55.4% Women; 39.8% Men; 4.8% not specified
Age: 18–71 (M = 26.8; SD = 12.7)
Country: England
|
Quantitative, randomized experimental study
|
Relationship between identification with a character, moral approval of their actions and its causal attribution
|
IPSAQ instrument (Kinderman & Bentall, 1996);
Identification Scale (Tal-Or & Cohen, 2010);
Ad-hoc Likert Scale on moral approval
|
5/5
|
Ten studies used questionnaires to evaluate parasocial relationships and identification (Table 1). Some of these instruments are the Parasocial Interaction Scale (Cale & Leets, 1999), Experience of Parasocial Interaction Scale (Hartmann & Goldhoorn, 2011) or the Identification Scale (Tal-Or & Cohen, 2010). Nevertheless, most studies (n = 7) used ad-hoc scales specifically designed for their respective research. These ad-hoc scales aimed at measuring variables related to parasocial relationships, identification, and character perception and traits. Two studies exclusively utilized such ad-hoc instruments (Ramasubramanian & Kornfield, 2012; Liebers & Schramm, 2017). In addition, various standardized questionnaires were used to assess attachment style (e.g. Attachment Style Questionnaire), interpersonal relationships (e.g. Experiences in Close Relationships-Revised Scale), personality (e.g. Big Five Inventory), and symptomatology (e.g. Center for Epidemiologic Studies Depression Scale).
Finally, the methodological quality of the studies was assessed using the MMAT, with scores ranging from three to five points. Most studies (n = 8) received a quality score of five points. The main methodological flaw identified was that some study samples could not be considered as representative of the target population. This issue was present in four of the articles (Schmid & Klimmt, 2011; Shedlosky-Shoemaker et al., 2014; Liebers & Schramm, 2017; Bernhold & Metzger, 2018).
The thematic analysis is presented below, organised according to the final version of the template. The template is divided into two main blocks: the first one integrates research on parasocial relationships with fictional characters and the second one on identification with fictional characters. Each of these blocks is divided into different sections that attempt to respond to the objectives of the systematic review.
Parasocial Relationships with Fictional Characters
The first content block of the thematic analysis analyses the characteristics of parasocial relationships, the factors involved in their development and the role that parasocial relationships play for individuals.
Characteristics of Parasocial Relationships
As previously explained, parasocial relationships, like social relationships, include platonic, but also romantic/sexual bonds. Karhulahti & Välisalo (2021) compile the terms fictophilia, fictosexuality and fictoromance to designate strong feelings of love or desire towards long-term fictional characters. What characterises fictophilia is what these authors call the fictophilic paradox: the contradiction between real feelings and the non-existence of the object of these feelings. It is this paradox that distinguishes these feelings from conditions such as erotomania, in which the distinction between reality and imagination is altered. This paradox, in fact, can be applied to any kind of parasocial relationship, whether romantic-sexual or not.
The nature of parasocial relationships makes it possible to establish relationships that provide a sense of security (Karhulahti & Välisalo, 2021). Parasocial relationships differ from social relationships with real people in that they do not carry the risk of harm or betrayal. This lack of risk provides a sense of control that is absent in interpersonal relationships (Karhulahti & Välisalo, 2021). However, it is important to note that this control is not absolute. Certain circumstances, such as the death of a character or the conclusion of a book series, game, or TV show, can affect parasocial relationships. In these situations, parasocial grief with fictional characters may demonstrate similarities to parasocial grief towards celebrities or even non-parasocial grief (DeGroot & Leith, 2018). Emotions such as grief, sadness, surprise, melancholy, or love are experienced and expressed. The character is remembered, and tributes are made in their honour.
Source Variables Involved in Parasocial Relationships
There are several factors when consuming a work of fiction that influence the development of parasocial relationships. The first one is mode of consumption. Liebers & Schramm's (2017) results indicate that ease of cognitive access is linked to a more pleasurable experience when consuming the story. Cognitive accessibility refers to the ease with which an individual can consume a fictional medium, such as reading, watching, or playing. It is determined by the level of effort required to concentrate on and comprehend the work. Therefore, the easier the story is to understand, the more likely it is to be enjoyed. This enjoyment, in turn, is related to greater transportation into the story and its world, which encourages parasocial interactions that can lead to parasocial relationships. Thus, there is seen to be a correlation between ease of consumption and the development of parasocial relationships.
As with people in social relationships, the characteristics of fictional characters influence the development of parasocial relationships. There seems to be a relationship between amicable parasocial relationship and perceived similarity to the character and its attractiveness, whether physical, mental, or social (Liebers & Schramm, 2017; Schmid & Klimmt, 2011). In contrast, in the case of romantic parasocial relationships, only physical attractiveness seems to be a direct predictor (Liebers & Schramm, 2017).
Consumer Variables Involved in Parasocial Relationships
Consumer also variables have been studied in relation to parasocial relationships. One line of research is related to attachment styles. Rain & Mar (2021) report that anxious attachment is related to stronger parasocial relationships, which would be explained by the proximity and intimacy seeking nature of anxious attachment. In contrast, avoidant attachment seems not to be related to parasocial relationships at all, which is consistent with the tendency of avoidantly attached people to keep their distance from others. Attachment style also appears to influence the type of characters with whom parasocial relationships are developed, as Rain & Mar (2021) also found evidence that anxious people tend to prefer sociotropic characters, i.e., those who prioritise the needs of others and are caring and protective. These types of characters may, through their personality, counterbalance the needs of the people with anxious attachment.
Schmid & Klimmt (2011) studied the influence of culture on parasocial relationships. Culture can be defined as the set of attributes shared by a set of human beings, including behavioural patterns, traditions, values, symbols, and beliefs (Dinkha et al., 2015). Culture influences identity and socialisation, making it virtually inseparable from the individual. In Schmidt & Klimt (2011)'s study, two types of cultures are distinguished: collectivist and individualistic. According to the definitions of Hofstede & Hofstede (2001) and Triandis & Suh (2002), collectivist cultures value family, interdependence, a sense of belonging and harmony, while individualistic cultures focus on the individual, independence and personal identity. Schmid & Klimmt (2011)'s findings seem to indicate that people from collectivist cultures develop stronger parasocial relationships than people from individualistic cultures.
Relationship with Identity and Personality
Parasocial relationships can become an important part of a person's identity, for example, through their own behaviour (Karhulahti & Välisalo, 2021). Even if these characters cannot reciprocate the relationship or interaction, individuals may engage in behaviours that counteract this lack, such as re-consuming the series or book from which the character originates or creating their own stories featuring the character (fanfiction).
Fiction allows access, through fantasy, to experiences that are not possible or desirable in the real world. It favours a safe exploration of relationships and fosters personal knowledge and growth (Karhulahti & Välisalo, 2021; Shedlosky-Shoemaker et al., 2014). In fact, parasocial relationships seem to have an impact on people's attitudes and behaviours, and may promote a change congruent with those presented by the character (Tian & Hoffner, 2010). This phenomenon seems to occur more strongly with fictional characters than with real-world celebrities or acquaintances, highlighting the power of parasocial relationships as a resource for self-concept formation (Shedlosky-Shoemaker et al., 2014).
Relationship with Mental Health
Evidence points to parasocial relationships with fictional characters as a useful coping mechanism for dealing with relationship anxiety or unmet needs for intimacy (Karhulahti & Välisalo, 2021; Rain & Mar, 2021). They are especially relevant for individuals experiencing other challenges in their lives, such as symptoms of depression. Individuals with anxious attachment or poor social relationships experience fewer depressive symptoms when engaging in parasocial relationships. However, in those with both anxious attachment and poor social relationships, parasocial relationships are linked to more depressive symptoms, not fewer (Bernhold & Metzger, 2018). Parasocial relationships can provide support in difficult situations, but they cannot fully compensate for affective-relational problems, although some people may seek refuge in them.
Similar findings have been made for avoidant attachment style. People with this type of attachment reported a positive correlation between hostile parasocial relationships (those towards characters who are hated) and depressive symptoms and loneliness (Bernhold, 2019). In this case, having good quality social relationships also function as a mitigating factor.
Stigma towards Parasocial Relationships with Fictional Characters
Parasocial relationships with fictional characters are often stigmatized and met with prejudice (DeGroot & Leith, 2018; Karhulahti & Välisalo, 2021). Therefore, discussions on parasocial relationships are more commonly found on the internet and social media, where individuals can seek support and resolve doubts while maintaining anonymity. Parasocial grief, in particular, is considered a disenfranchised grief (DeGroot & Leith, 2018). The experience of parasocial grief is often stigmatized and judged by others, despite its validity and justification. Additionally, the nature of parasocial relationships and the use of social networks can contribute to maladaptive cyclical grief (DeGroot & Leith, 2018).
Identification with Fictional Characters
The second content block of the thematic analysis, dedicated to identification with fictional characters, analyses the factors involved in this process and the role it plays for individuals.
Source Variables Involved in Identification
As with parasocial relationships, there are variables related to the mode of consumption of the story that favour identification processes with characters. The most important is transportation into the story, which is characterised by greater cognitive attention to the narrative, more vivid mental imagery, and more intense emotional responses to the events (Sestir & Green, 2010; Shedlosky-Shoemaker et al., 2014).
Character characteristics are also relevant in identification processes. Sestir & Green (2010) point out that character characteristics can promote identification by two means. The first one occurs when consumers detect a real or perceived similarity between the fictional character and themselves. The second one is due to the consumers’ identification of desirable characteristics in the character (wishful identification).
Consumer Variables Involved in Identification
Age appears to be negatively related to identification, wishful identification, and imitation of fictional characters' behaviour (Ramasubramanian & Kornfield, 2012). In terms of gender, women report more identification and wishful identification than men to fictional characters (Ramasubramanian & Kornfield, 2012; Shedlosky-Shoemaker et al., 2014).
Attachment style have also been studied in identification processes. Complementary to the role of anxious attachment as a predictor of parasocial relationships, studies have found that avoidant attachment appears to positively correlate with identification, especially with independent and autonomous characters (Rain & Mar, 2021). These are traits that an avoidant person is likely to find desirable, which favours identification processes.
Moral approval of a character's actions also seems to be related to identification processes. Teasdale et al. (2021) found a positive correlation between the two variables. However, whether moral approval promotes identification or identification promotes moral approval is difficult to determine.
Culture influences the way characters are both constructed and understood (Ramasubramanian & Kornfield, 2012; Schmid & Klimmt, 2011), so it is safe to assume that it will also be related to identification processes. Different cultures develop different types of fictional characters, depending on the context of their creation (Ramasubramanian & Kornfield, 2012). Thus, collectivist cultures tend to create collectivist characters, while individualist cultures tend to create individualist characters. When people consume a work of fiction from another culture, they are interacting with stories and characters created by and for that culture, not their own. This would make them more appealing to certain audiences, affecting the relationships consumers stablish with characters (Ramasubramanian & Kornfield, 2012).
Relationship with Identity and Personality
Several studies indicate that a high level of perceived similarity and identification favours changes in self-perception congruent with the character characteristics (Sestir & Green, 2010; Shedlosky-Shoemaker et al., 2014; Tian & Hoffner, 2010). However, it appears that wishful identification has these same effects to a greater degree (Shedlosky-Shoemaker et al., 2014; Tian & Hoffner, 2010). Wishful identification is also related with direct and conscious imitation of character behaviour (Ramasubramanian & Kornfield, 2012).
Development of an Explanatory Model: Relationships between Parasocial Relationships and Identification with Fictional Characters
As a result of this systematic review, an explanatory model has been developed. This model brings together the scientific research reviewed and makes explicit the links between the different elements. A schematic of the model is shown in Fig. 3.
The model shows that parasocial relationships and identification are not independent processes, but influence each other. According to research, parasocial relationships seem to have more power to influence identity than identification alone, the latter being a good predictor of parasocial relationships (Tian & Hoffner, 2010). It seems that identification, which is more immediate and occurs while consuming the story, predisposes to the later development of a parasocial relationship with the character.
Throughout the review, several links have been identified between consumer and source variables, and the relationship of these processes to personal identity. This suggests that these are not factors that independently influence the processes of identification and parasocial relationships. These connections between variables form a more complex net of interdependent phenomena. As shown in the graphical representation of the model, consumer variables interact with source variables to make people more or less sensitive to characters with certain characteristics. These characteristics influence personal identity, through both parasocial relationships and identification processes.