Self-knowledge is a central issue in adolescent development [1
]. Self-knowledge can be considered as a personal theory (or theories) that a person construes based on his/her experience, so as to be able to anticipate the events of his/her world, to inspire him/her with appropriate (adaptive) actions in each moment, as well as to maximize psychological well-being (e.g., [3
]). The transformations of adolescence, like body changes (e.g., [7
]), new social contexts and expectations [3
], and cognitive advances in adolescent lives [10
], press for changes in both the content and structure of self-knowledge [5
]. Thus, adolescents can count on new guides to make sense of a diversity of events and inspire adaptive and autonomous behaviors in a multitude of situations.
Many constructs have been proposed to describe the structure and organization of self-knowledge in the context of adolescence by approaching, in different ways, the issues of differentiation, integration, and/or, less often, internal conflicts. This research line has provided important evidence and theoretical advances, but it still faces some challenges. First, studies that include several of these constructs of self-structure are scarce (except for self-esteem). Second, the constructs (and measures) utilized across the literature vary greatly, making them difficult to compare to each other, thus challenging or hampering researchers and practitioners’ efforts to obtain a comprehensive understanding of their role in the organization of self-knowledge (e.g., [12
]). Third, some authors defend the superiority of idiographic data to assess self-structure, but measures based on (self-) descriptions provided by researchers prevail [12
]. Finally, certain variables considered central in theory have been underrepresented in research, as it is the case of internal conflicts. Therefore, this study addresses some of these limitations by exploring self-knowledge organization in adolescence using George Kelly’s personal construct psychology (PCP) [6
] as a comprehensive framework, both theoretically and methodologically (see also [15
1.1. The Perspective of Personal Construct Psychology on Self-Knowledge
From the beginning of life, a person lives immersed in a succession of events, objects, other people, as well as the experience of oneself. PCP [6
] proposes that, based on the similarities and differences perceived between events, one builds bipolar dimensions of meaning (the personal constructs; for instance, “quiet–communicative”) and, in this way, makes life more predictable. The self and significant others are the most challenging and crucial elements to make sense of for an individual’s adaptation to the contexts, mainly social, we live in. So, construing self and others is the core psychological activity for adaptation, according to Kelly.
The self is, when considered in the appropriate context, a proper concept or construct. It refers to a group of events which are alike in a certain way and, in that same way, different from other events. The way in which the events are alike is the self. That also makes the self an individual, differentiated from other individuals [6
] (p. 131).
The self is thus the pole of a “self–no-self” bipolar construct (or, even better, a “self–others” construct), “which in turn is construed” [16
] (p. 456). Therefore, self-knowledge can be defined as a system of personal constructs conceived by the person from the similarities and differences they perceive between other people and themselves, and that is useful for the person to anticipate events and guide successful action in the interpersonal world (see also [17
]). Thus, in this theory, knowledge about the self is entangled with knowledge about significant others, as proposed by other theories of the self [18
] which has also been supported by cognitive neuroscience [20
]. Furthermore, people construe a multiplicity of selves or a community of selves [21
]. In addition to the Actual Self, people construe, for instance, an Ideal Self, selves as perceived by various others, or selves in particular contexts, relationships, or roles (see also [22
Moreover, Kelly’s theory addresses how these personal theories change according to the flow of experience. In order to prevail, personal constructs need to be validated against experience. When a personal construct system is not able to anticipate events adequately, invalidation occurs in a process that activates unpleasant emotional experience [16
]. Then, some degree of revision of the current self-system is needed.
According to Kelly, constructs are not isolated units of meaning. Rather, they relate to other constructs in a hierarchical network where some constructs are superordinate and subsume others, the subordinate constructs. At the top of the system, core constructs are more stable and resistant to change than subordinate or peripheral constructs as they assure a sense of continuity and personal identity. Otherwise, their invalidation would render invalid large portions of the self-system and compromise a person’s ability to predict events. In addition, the structure of personal construct systems is proposed to increase in differentiation and hierarchical integration as life goes on, thus becoming more able and flexible in predicting a larger array of personal and interpersonal events.
In sum, in a certain sense, PCP also comprises a developmental point of view, although different from a normative developmental approach. Yet this perspective can be useful to understand change in self-knowledge organization in the context of daily life events and emotional experience. Despite often being regarded as a psychology of adults, PCP addresses human-construing processes in general and does not exclude children and adolescents. Indeed, in his masterwork, Kelly [6
] profusely uses examples of a child’s construing. Jablonsky and Lester [23
] noticed the scarcity of developmental studies (cross-sectional or longitudinal) in the PCP field, especially those concerning changes throughout adolescence and adulthood.
To assess the personal construct systems, Kelly [6
] conceived the repertory grid technique (RGT), a structured procedure similar to an interview (see [24
]). This tool eloquently illustrates PCP’s perspective on self-knowledge. First, interviewees name a sample of significant others and self-elements (e.g., most frequently Actual Self and Ideal Self, but other self-elements can be considered, like “Ought Self” or “Self as perceived by particular others”), then they are asked to identify similarities and differences between them; in this way, a sample of bipolar constructs, like “sad–happy”, is elicited. Finally, individuals rate every self and other element in each of the elicited constructs (usually on a 7-point Likert scale). At the end, this technique provides a qualitative and quantitative data matrix (see Figure 1
) that allows the computing of several cognitive structure measures, such as differentiation, polarization, discrepancies between Actual Self, Ideal Self, and Others (as perceived by the interviewee), as well as several measures of psychological conflict [24
Compared to more widely used instruments, the RGT provides idiographic self-knowledge (concerning content and organization patterns) and allows qualitative and quantitative standardized measures, a recommended combination to assess self-knowledge organization [12
Finally, previous studies have revealed the RGT as a promising technique for understanding several psychological aspects central to adolescents’ self-knowledge development, which will be reviewed in the second part of the next two sections, preceded by a summary of contributions from developmental approaches.
1.2. Differentiation and Conflicts in the Organization of Self-Knowledge in Adolescence
An influential and comprehensive perspective on the development of the self in adolescence is offered by S. Harter ([5
], for a review). According to her perspective, important cognitive changes occur in adolescence that, combined with new social contexts, promote reorganizations of self-knowledge. S. Harter describes a detailed developmental trajectory comprising three phases across adolescence. During early adolescence, childhood self-representations give place to an increasing number of single self-descriptive abstractions. Adolescents personal characteristics increasingly differentiate in terms of relational contexts (e.g., me-with-my-father, me-with-my-best-friend). Similarly, early adolescents distinguish their competence in an increasing number of different life domains (e.g., scholastic competence, social competence, physical appearance, behavioral conduct, close friendship, romantic appeal, job competence). At the same time, self-esteem becomes more differentiated according to different social contexts. By middle adolescence, adolescents perceive contradictions among the differentiated personal characteristics; thus, they experience psychological conflicts and preoccupations about the existence of false selves and the identification of their true self (e.g., “talkative” or “secretive”?) [5
]. Later, those attributes formerly perceived as contradictory can be combined into a more coherent self-picture (e.g., “Basically, I am flexible: I am talkative with friends and secretive with my parents”). Research suggests that older adolescents are often able to organize their self-knowledge in this integrative way, provided they can count on the appropriate support (e.g., a more mature person) [5
]. Other studies suggest that some kinds of conflicts diminish at late adolescence, but others do not [28
]. Also, older adolescents become capable of a more nuanced and less polarized view of themselves and others. Indeed, external support seems to have an important role in the adolescent task of organizing self-knowledge. The developmental sequence described by S. Harter [5
] is particularly evident when adolescents receive some support in the task of organizing self-knowledge (e.g., identifying conflicts); when support is not available, adolescents show less cognitive sophistication and developmental advances seem to be slower and more linear [27
According to the PCP’s framework, self-knowledge differentiation and hierarchical integration increase as life goes on, and adolescence should not be an exception. Considering that the RGT has been fruitful in providing measures of differentiation, integration, and conflicts, in this study, we aim to measure two of the most consolidated aspects of the overall cognitive structure (and its sophistication), namely, global differentiation and polarization [30
]. Differentiation refers to the diversity of different dimensions of meaning that a person has available to anticipate interpersonal and personal events. Polarization refers to rigid (black and white) knowledge about self and others along the dimensions of meaning, as opposed to the ability to build up a nuanced perspective.
Reviewing studies about developmental change in children and early adolescents, Jablonsky and Lester [23
] suggested the predicted developmental trend of increasing differentiation and integration. Also, they mentioned an increase in the number of constructs and the constructs becoming more abstract and less concrete (also [28
]). However, only a few developmental studies (cross-sectional or longitudinal) have addressed this issue at different stages of adolescence. Zhang et al. [14
] found no correlation between differentiation and age across adolescence in a sample of Chinese adolescents. However, they attributed this result to the nomothetic approach they used (the constructs were provided by the researchers). Other studies measured differentiation and polarization in adolescence but included adolescents of only one age group [31
Yet, in a system where constructs are considered to be interconnected, it is possible that contradictions arise between constructs and imperfections happen in the integration of the system, like fragmentation and conflicts. The RGT has been fruitful in providing measures of intrapersonal conflict (for a review see, for instance, [26
]). In a cross-sequential study involving participants from 6 to 18 years old, Oosterwegel and Oppenheimer [28
] conceived a composite measure of conflicts within (e.g., Ideal Self from perceived parents’ perspective) and between self-concepts (e.g., Ideal Self and Actual Self from own perspective), partially based on RGT data. They found that the number of conflicts within self-concepts (e.g., Actual Self from adolescent’s own standpoint) peaked between 12 and 14 years old and stabilized in subsequent years [33
]. Concerning conflicts between self-concepts, they found different pathways. The conflicts between Actual and Ideal Selves from the perceived perspective of parents increased throughout childhood, stabilized during adolescence, and tended to decrease slightly in late adolescence. However, they found no age-related changes in the conflicts between other self-concepts, such as the conflict involving Actual and Ideal Selves from the youth perspective [28
]. In sum, they found that different conflicts showed different developmental pathways across childhood and adolescence, some possibly conforming to the developmental sequence proposed by S. Harter [5
], but others not.
However, in Oosterwegel and Oppenheimer studies, as in S. Harter’s [29
], the identification of conflicts strongly relies on adolescents’ awareness, and a certain amount of support (especially in S. Harter’s research) was provided for adolescents to relate their self-descriptions. Conversely, RGT offers the possibility of computing several measures of conflict that do not require the awareness of the person nor explicit cognitive activity in identifying conflicts (as most likely happens when adolescents live their daily lives).
In this study, we focus on implicative dilemmas [26
], a particular kind of conflict identified in repertory grids (described in detail in the Methods section). Briefly, implicative dilemmas involve the Actual and Ideal Selves as they are perceived by two distinct personal constructs that are linked. This is a conflict in which a desired change in one construct is blocked because it implies an undesired change in another attribute. Research has shown that this type of conflict is related to psychological distress [34
], including in adolescents [31
]. However, no previous studies have focused on developmental differences in this kind of conflict.
1.3. Self-Discrepancies and Identity Construing in Adolescence
In the developmental literature, another piece of evidence about cognitive differentiation refers to the increasing distance between the Actual Self and the Ideal Self throughout adolescence [36
]. This Actual–Ideal self-differentiation (A–I; also called self-disparity or self-discrepancy by different authors), similar to other kinds of self-differentiation, has been associated with different, sometimes contradictory, psychological meanings. For instance, the increase of self-discrepancies with age has been conceived as part of the internalization of self-guides taking place during adolescence [37
]. During this process, external self-evaluation and self-regulation standards provided by significant others (as models, or as young people perceive others’ expectations to the self) give place to adolescents’ own standards and to a more autonomous, independent self-regulation. In this context, the research on self-discrepancies suggests that the Ideal Self of adolescents seems to become progressively more differentiated from the significant others (as they are perceived by the adolescent), from the Ideal Self he/she perceives others have for him/her, and from the self-perceived Actual Self, as they age [5
In addition, the A–I differentiation has been considered by some authors to be associated with self-esteem, an association especially important in the context of the adolescents’ development [5
]. According to this perspective, self-esteem emerges from the distance between the actual attributes one perceives about him or herself and those which he/she aspires to possess. Some evidence supports this association between A–I and global self-esteem, as measured by traditional standardized scales [40
], especially when idiographic measures of Actual and Ideal Self attributes are considered [43
]. However, evidence about changes in global self-esteem from childhood on is not consensual and is not always consistent with the developmental increase proposed to A–I throughout adolescence. Some studies suggest that global self-esteem diminishes throughout adolescence to start increasing in emerging adulthood [44
], whilst others propose that it diminishes only in early adolescence and increases in the following years [46
The research on individuation and identity formation is another important contribution to the field [3
]. Adolescents are expected to become progressively more independent from parents, to define autonomous relationships with the world and to establish themselves as a singular constellation of attributes. Accordingly, there is evidence that a sense of being different from others increases across adolescence [49
]. However, in the end, as young people go through emerging adulthood, the role of integrative processes seems to be of great importance in the task of combining a diversity of psychosocial possibilities in a unique, committed identity [3
In the PCP field, the topics of identity and self-discrepancies are linked. Different authors have proposed, based on RGT, that the construing of self-identity relies on the degree of similarity between the self-perceived Actual Self, how significant others are perceived, and the Ideal Self [24
]. First, the discrepancy between the Actual Self and significant others (as they are perceived by the adolescent) (i.e., A–O) is considered a measure of identification with others, or distinctiveness, an important component of the classic view of identity [52
]. Adams-Webber (see, for example, [54
]) argued that people tend to perceive similarities between themselves and others in about 62% of characteristics and that this balance is achieved in adolescence. From his point of view, the supremacy of characteristics shared with others (as opposed to those that are distinct) serves the function of better differentiating self and others, such as a figure (the differences) against a background (the similarities). Some developmental research within the scope of PCP shows, for example, that children (8–11 years) perceive themselves to differ from their parents less than adolescents (see [54
]), and that this differentiation increases throughout adolescence. Yet adolescent efforts to maximize this distinction can compromise the validation of the self [55
Second, the differentiation between Actual Self and Ideal Self can be understood as a degree of acceptance of the self and/or as a predictor of self-esteem [56
]. A high discrepancy promotes invalidation of the self, unpleasant effects, and a need to change [55
]. Results regarding the A–I developmental pathway are inconclusive, as some studies showed no changes across childhood and adolescence [28
] and others suggested a peak in middle adolescence [55
]. Less studied, the differentiation between Ideal Self and Others (I–O) refers to the appreciation that one has about Others as more or less adequate [24
] or as potential role models, thus inspiring the direction of personal change [55
]. This distance increases across adolescence, and greater distances would hinder the social validation of the adolescent’s self [55
Overall, these three self-discrepancies have been studied more in relation to psychological adjustment than to cognitive development. Previous studies on adults suggest that greater differentiations are related to psychological distress [56
]. A study on adolescents found a relation with depressive syndrome only for the A–I differentiation [31
]. In addition, with the inclusion of the Ideal Self and the anticipation of a possible (maybe future) self it represents, PCP somehow addresses another important component of identity, the sense of personal continuity [52
1.4. The Current Study
In summary, different developmental approaches suggest an increase of self-knowledge differentiation across adolescence, with very few studies highlighting a peak of self-perceived conflicts within self-knowledge by middle to late adolescence. These conflicts could possibly be reorganized in a more coherent, integrative way by late adolescence or emerging adulthood since adolescents are provided with support. In addition, intrapersonal conflicts seem to be understudied despite the numerous theories addressing its role in development and psychological adaptation (for some exceptions, see [5
]). In our view, PCP could offer an integrative approach, providing potential insights into the understanding of self-knowledge development during adolescence, thus addressing a gap in the current literature.
The aim of this cross-sectional study is to explore the self-knowledge organization in adolescence with the help of PCP’s conceptual and methodological contributions when considering predictions from developmental perspectives. Among PCP’s conceptual contributions, we highlight the entanglement of knowledge about the self and knowledge about significant others or its ability to understand personal change in the face of the experience. Examples of methodological contributions are a set of structural, idiographic measures derived from RGT, including conflicts, which may reveal cognitive structures that people are not immediately aware of.
As mentioned above, the administration of the RGT does not provide support to adolescents in the particular task of relating self-attributes. Hence, developmental differences are expected to be slow to appear; for instance, the emergence of conflicts and, especially, their subsequent resolution, would be expected to occur later than proposed by S. Harter’s developmental sequence [5
]. With this in mind, this study compares two periods of adolescence, namely, early and late adolescence.
Our central hypothesis anticipates that self-knowledge organization will be different in early and late adolescents, such that the older sample will show (a) higher global differentiation; (b) lower polarization; (c) more participants with intrapersonal conflicts; and (d) higher discrepancies between Actual Self, Ideal Self, and Others.