The concept of social representation traces back to Serge Moscovici [5], who developed the theory in his seminal work on the image of psychoanalysis in French society. Drawing on the Durkheimian notion of collective representations [12], Moscovici defines social representations as systems of dynamic, socially constructed knowledge that fulfill practical functions in everyday life: they allow individuals to orient themselves within their material and social environment and make communication possible among members of a group [5] [13].

Jodelet [7] offers a widely cited definition of social representation as a “form of knowledge, socially elaborated and shared, having a practical aim and contributing to the construction of a common reality for a social group.” What stands out in this definition is the emphasis on three interconnected elements: how this knowledge is socially constructed, the fact that it is shared among peers, and its practical, everyday purpose.

The structural approach, initiated by Abric [6] [8] [14], posits that representations are organized around a central core that gives them meaning and determines the relationships between their elements. The central core is characterized by its stability, resistance to change, and connection to the collective memory and norms of the group [6] [10]. Peripheral elements are more flexible: they allow the representation to adapt to contextual variations while protecting the central core [6].

Social representations fulfill four functions [6]: knowledge functions, which allow for the understanding and explanation of reality; identity functions, which situate individuals and groups within the social field; orientation functions, which guide behaviors and practices; and justificatory functions, which allow for the a posteriori justification of stances and conducts.

Academic cheating encompasses a set of behaviors by which a student seeks to gain an advantage in an assessment without respecting the rules applicable to all: copying from a peer, communicating without authorization during examinations, using prohibited materials (“cheat sheets,” or “badjô” in the Ivorian context), or plagiarizing [15] [16].

From a sociological perspective, cheating constitutes a form of deviance. Becker’s [17] interactionist approach emphasizes that deviance is not an intrinsic property of the act, but the consequence of the application of judgments and sanctions by social groups. This perspective is relevant to academic cheating: students may not perceive their behavior as deviant within their peer group, even when it violates institutional rules.

Empirical research has identified several factors associated with cheating. Jensen, Arnett, Feldman, and Cauffman [18] show that students often judge cheating as less serious than adults do and perceive it as a common practice, with the acceptability they grant it being linked to their own cheating behaviors. McCabe, Treviño, and Butterfield [2] highlight the determining role of contextual factors, particularly peer norms, in the frequency of academic dishonesty. Magogeat [16], giving voice to cheating high school students, highlights the practical logics and justifications that accompany these behaviors.

This study relies on two complementary frameworks. Moscovici’s theory of social representations [5] provides the conceptual foundation for understanding how students collectively construct the meaning of cheating: representations are neither purely cognitive nor purely social but emerge from the interaction between psychological processes and social contexts [13]. Abric’s central core theory [6] [8] allows for the identification of its organizing principles: the core, composed of a limited number of elements, determines the meaning of the representation and its stability [10].

Social representation can be understood as a system of knowledge that is collectively shaped and shared within a group, primarily serving a practical function by helping members construct a common reality [7]. As such, it sits at the delicate interface between the psychological and the social [11] [19].

Academic Cheating. The set of behaviors by which a student seeks to gain an advantage in an assessment by violating the rules applicable to all, notably by copying, communicating without authorization, or using prohibited materials [15] [16].

Deviance. A conduct that deviates from established norms, and whose qualification depends on social judgments rather than on the intrinsic properties of the act [17].

Representation: From the Latin repraesentatio (“action of placing before the eyes”), representation refers to the process by which an image is presented to the senses or an image formed in the mind of an absent object or concept [20].

Social Representation: A form of socially elaborated and shared knowledge with practical aims, contributing to the construction of a common reality for a social group [7] [13]. Social representations are situated at the interface of the psychological and the social, simultaneously cognitive and social in nature [11].

Academic Cheating: An act by which a student obtains advantages or rewards without respecting applicable rules, including copying from others, unauthorized communication, and use of prohibited materials [21]. It constitutes a form of deviance [17] characterized by the transgression of institutional norms for personal gain.

Deviance: An act or behavior departing from established norms, whose characterization depends on social judgments rather than inherent properties of the act itself [17].

To achieve our objectives, we employed a descriptive cross-sectional design. This approach is grounded in the structural study of social representations and relies on the hierarchical evocation technique and prototypical analysis, which allow for the identification of the constitutive elements of a representation and the formulation of hypotheses regarding its organization [9] [22].

The research was conducted at Lycée Martin Luther King de Bouaké (LMLKB), a public secondary school located in the Kennedy neighborhood of Bouaké, Côte d’Ivoire. Created in 1983 as a private institution before being transferred to state management, it accommodates approximately 1990 students from the sixth grade to the terminal year, with a teaching staff of 92.

The target population consisted of Grade 9 students at LMLKB. Through convenience sampling, 77 students were selected-48 boys and 29 girls-distributed across six Grade 9 classes.

Data were collected using a hierarchical evocation questionnaire based on the free association technique [22]. Participants received the inductive word “cheating” and were asked to write at least four words or expressions that came to mind. Next, participants ranked these evocations from most to least important (1 = most important, 2 = second most important, and so on). This technique enables simultaneous analysis of the categorical system, content, and organization of the representation [9].

We collected the data during the 2023-2024 academic year. With the teachers’ cooperation, we administered the questionnaires in class, ensuring a quiet environment that allowed students to answer thoughtfully. Before they began, we explained the study’s objectives to the participants and guaranteed the anonymity and confidentiality of their responses.

Data were analyzed using EVOC 2005 software [9], which performs prototypical analysis by cross-referencing the frequency and rank of evocations. The resulting four-cell table distinguishes central core elements (high frequency, low mean rank) from peripheral elements (low frequency, high mean rank) and potential change elements. Following Vergès’ [9] procedure, elements were categorized based on a frequency threshold (minimum of 10 evocations for the general sample; adjusted for gender subgroups) and a mean rank threshold (2.5).

To compare genders, we used Pearson’s chi-square tests (without continuity correction) alongside Fisher’s exact test. (for expected frequencies < 5) were conducted, along with the calculation of the phi (φ) coefficient to assess the strength of associations. A Bonferroni correction was applied to control for multiple comparisons (α = 0.0125 for four comparisons).

The 77 participants generated 317 evocations, comprising 130 distinct words, averaging 4.12 evocations per participant. However, given that this is a free evocation task, the mean rank of an item depends solely on the terms produced by the respondents who cited it. Table 1 presents the distribution of the main elements according to their frequency of occurrence and mean rank.

Table 1.Prototypical analysis of the social representation of cheating (N = 77). This option adds explicit words like “Frequency” and “Low/High” to make the table instantly readable.

Note. f = frequency of evocation; m = mean rank of importance (1 = most important). “Badjô” refers to a cheat sheet in Nouchi (Ivorian urban slang).

The central core zone comprises four elements that are both frequent and judged important: “stealing,” “lying,” “fraud,” and “cheating.” These elements constitute the most probable candidates for the status of central elements of the representation [9] [22]. The first periphery includes “badjô,” a frequently cited but lower-priority term. The contrast zone contains “communicating,” evoked by few participants but ranked among the first; this type of element may signal a subset of students whose representation is more focused on the practical modalities of cheating. The second periphery groups “looking aside,” “corruption,” “scamming,” and “deception,” which are more individual elements and more dependent on context.

According to the prototypical analysis, the items “stealing,” “lying,” “fraud,” and “cheating” are particularly salient in this representational study.

However, the item “stealing,” with the highest frequency (35) and a mean rank of 2.26, appears as the most shared element. As for cheating, it is conceived as an undue appropriation consisting of obtaining grades, diplomas, and ultimately opportunities that are not the fruit of one’s own work.

Regarding the item “Lying,” with a high frequency (f = 32; m = 2.44), this evocation presents cheating as a deception towards all school stakeholders and more specifically, teachers, parents, peers, and oneself regarding one’s actual knowledge.

Concerning the item “fraud,” with a mean rank of 1.81, this central zone element, also judged important, refers to the regulatory dimension of cheating as a violation of established rules.

The item “Cheating,” the verbal form of the inductive word, also appears in this zone (f = 11; m = 2.00). Its presence likely reflects a restatement of the inductive word rather than substantive content. As the verbal form of the inducer, this item very likely stems from a lexical artifact rather than a substantive component of the representation, thereby undermining its semantic content.

The central core zone of the representation of cheating among Grade 9 students at LMLKB is essentially organized around three substantive contents: stealing, lying, and fraud. Together, they construct cheating as a morally negative act, associating the appropriation of unmerited advantages, deception, and violation of institutional rules.

Regarding the peripheral system, it refers to a more concrete dimension. The item “Badjô,” derived from Nouchi, designates the instrument of cheating. This term bears witness to the linguistic creativity of students and suggests a certain normalization of these practices within peer culture. “Looking aside” and “communicating” refer to the practical modalities of cheating during an examination.

Girls produced 113 evocations (60 distinct words, i.e., 3.90 per student) and boys 204 evocations (99 distinct words, i.e., 4.25 per student). Table 2 presents, for each gender, the frequency, the proportion of students concerned, and the mean rank of the central core zone elements for the entire sample of respondents.

Among the girls, the three substantive elements have a mean rank below the 2.5 threshold and therefore fall within the central zone. “Lying” is by far the most shared element (62.1% of girls; m = 2.22), followed by “fraud” (37.9%; m = 1.91). “Stealing” is evoked by slightly more than a quarter of the girls (27.6%), but with the lowest mean rank in the entire study (m = 1.38): the girls who associate cheating with stealing place it almost systematically at the top of their evocations. The girls’ representation is thus dominated by the moral dimension of lying.

Among the boys, “stealing” is the most shared element (56.3% of boys), but its mean rank (2.52) is just above the 2.5 threshold; it therefore belongs to the first periphery rather than the central zone. The same applies to “lying” (29.2%; m = 2.71). The boys’ central zone consists of “fraud” (31.3%; m = 1.73) and “cheating” (20.8%; m = 2.10). The boys’ representation thus associates a primary anchoring in fraud, i.e., the violation of rules, and a very widespread but less prioritized evocation of stealing. Table 3 compares the proportion of girls and boys who evoked each of these elements.

Table 2.Elements of the central core zone by gender.

Note. f = number of students who evoked the term; % = proportion of students in the subgroup; mean rank = mean rank of importance (1 = most important). The number of “cheating” evocations among girls is deduced from the total numbers (11 - 10); its mean rank is not available.

Table 3.Comparison of evocation proportions by gender.

Note. Statistical comparisons were conducted using Pearson’s chi-square test (without continuity correction) and the two-sided Fisher’s exact test. φ = association coefficient. Bonferroni-adjusted threshold for four comparisons: α = 0.0125.

The difference is significant for the item “lying,” evoked twice as often by girls (p = 0.005), and remains so after Bonferroni correction; the effect size is moderate (φ = 0.32). The item “Stealing” is evoked significantly more often by boys the 0.05 threshold (p = 0.014), but this difference becomes borderline after correction (α = 0.0125). The observed gap for the item “cheating” (p = 0.035; Fisher p = 0.045) does not withstand the correction and relies on small numbers among girls. Finally, the item “fraud” is evoked in comparable proportions by both genders (p = 0.548) and appears in the central zone of both groups: it is the most consensually shared element.

The central core zone of the representation of cheating among Grade 9 students at LMLKB is organized around three substantive contents: stealing, lying, and fraud. Together, they characterize cheating as a morally negative act, associating the appropriation of unmerited advantages, deception, and violation of institutional rules.

The peripheral system refers to a more concrete dimension. The item “Badjô” designates the instrument of cheating. Derived from Nouchi, this term bears witness to the linguistic creativity of students and suggests a certain normalization of these practices within peer culture. “Looking aside” and “communicating” refer to the practical modalities of cheating during an examination.

Intra-group analysis shows that girls and boys share a common overarching representation, centered on fraud and morally negative, but its components do not have the same salience. Among girls, cheating is primarily a lie, i.e., an attack on honesty and trust; among boys, it is primarily a fraud, and very often associated with stealing, i.e., obtaining an unmerited advantage. These differences pertain to the relative weight of the elements within a shared representation rather than to two distinct representations.

Our findings reveal that cheating is fundamentally a socially constructed reality. More importantly, the most widely shared understanding of it among the students is deeply moral in nature. The central system, composed of “stealing,” “lying,” and “fraud,” reflects a fundamentally negative evaluation of cheating, echoing the Durkheimian conception of collective representations as an expression of shared moral sentiments [12] and the conception of social representations as systems of values and practices [5].

The fact that “stealing” is the most frequent element and “fraud” the most important suggests a dual anchoring of cheating in moral transgression (taking what belongs to others) and rule violation (circumventing institutional regulations). This dual anchoring is consistent with the role the structural approach attributes to the central core, as both a generator of meaning and an organizer of the representation [6].

The coexistence of a disapproving central zone and peripheral elements linked to concrete practices (composed of “badjô” and “looking aside”) illustrates the protective function of the peripheral system described by Abric [6] and Flament [10]: students can condemn cheating in principle while acknowledging its routine presence in their daily school life.

Gender differentiation nuances this observation. The fact that girls evoke lying more (moral and relational dimension) and boys evoke stealing more (transactional dimension) is compatible with research on differentiated moral socialization, according to which girls are more oriented toward an ethic of relationship and trust [23], and with Nucci’s [24] analyses on the diversity of domains of moral judgment. This interpretation remains, however, a hypothesis: the present data do not measure socialization and do not allow for establishing the origin of these differences. The sharing of fraud by both genders, at a priority rank, conversely reinforces the hypothesis of its centrality in the common representation.

These results align with those of Jensen, Arnett, Feldman, and Cauffman [18], according to whom students recognize the reprehensible nature of cheating while perceiving it as a common practice. The emergence of the item “badjô” as a peripheral element illustrates this normalization. They also echo the work of Magogeat [16], which shows that cheating high school students inscribe their practices within situational logics and shared justifications. This aspect refers to the work of McCabe, Treviño, and Butterfield [2], highlighting the weight of peer group norms. In the African context, these results extend Bipoupout’s [1] findings on the scale and stakes of cheating in African school systems.

These results require the training and vigilance of teachers. The elements “looking aside” and “communicating” point to the opportunities offered by the examination situation. The surveyed students seem to advocate for increased teacher vigilance and spatial arrangements such as seating plans [16].

It is furthermore necessary to design assessments that require the student’s contribution. The prominence of the item “stealing” suggests that students conceive of cheating as the appropriation of an unmerited result. For this, favoring assessments that emphasize understanding and application rather than simple recall can reduce the interest in and opportunities for cheating.

However, “Integrity Education” should be prioritized. The moral dimension of the representation (lying, fraud) indicates that students recognize the transgression inherent in cheating. This suggests a receptivity to ethical education programs valuing integrity and honest success. Teachers, administrative staff, and parents must be associated with this educational project.

The “gender” dimension deserves serious consideration. Consequently, gender must be taken into account in various interventions. The observed differences suggest adapting the messages: for girls, emphasize the relational consequences of cheating (loss of trust, damage to reputation). For boys, insist on the illegitimacy of the advantages thus obtained. The common foundation-fraud as a rule violation-can serve as a shared message for all students. The effectiveness of this differentiation must be evaluated before any generalization.

This study has several limitations that warrant consideration. First, because the research was confined to a single institution, we must be cautious about generalizing the findings to other contexts. Second, the sample size is relatively small and was obtained through convenience sampling, which may have introduced selection bias. Additionally, the underrepresentation of girls in the sample (29 compared to 48 boys) limits the precision of our estimates for this specific subgroup. Third, the data are self-reported and may be affected by a social desirability bias. Fourth, prototypical analysis allows for the formulation of hypotheses on the centrality of elements but does not demonstrate it: the central status of the items “stealing,” “lying,” and “fraud” must be confirmed by complementary techniques, such as the mise en cause (putting into question) or the centrality questionnaire [22]. Fifth, the distribution of elements depends on the thresholds retained; items close to the mean rank threshold, such as “stealing” among boys (2.52), “badjô,” or “corruption,” must be interpreted with caution. Finally, gender comparisons rely on multiple tests applied to modest-sized subgroups, which increases the risk of a Type I error; only the difference relating to “lying” withstands the Bonferroni correction.

Future research could: 1) verify the centrality of the identified elements, for the entire sample and for each gender, through centrality tests; 2) extend the study to other institutions and school levels in Côte d’Ivoire, with balanced numbers of girls and boys; 3) examine the socialization processes likely to explain the observed gender differences; 4) use qualitative interviews to explore the meanings associated with the evocations; 5) examine the link between representations and actual cheating behaviors.