Validation of French Versions of the Internal and External Sexual Consent Scales: Associations with Intimate Relationship and Sociodemographic Variables

1. Introduction

Sexual consent is a central yet complex construct in the study of intimate relationships, carrying direct implications for the prevention of sexual violence (Gilles-Noguès et al., 2025). Despite its apparent simplicity in everyday discourse, consent resists straightforward definition, and the conceptual ambiguity surrounding it has significant consequences for both legal frameworks and prevention practice (Fenner, 2017). Charlene Muehlenhard and her colleagues (2016) propose three conceptualizations. The first defines consent as an internal state of willingness: a subjective psychological experience of desire, readiness, and agreement. This definition captures what consent fundamentally is, but it remains largely inaccessible to outside observers, including legal systems, which cannot directly access another person’s mental states. The second conceptualizes consent as an explicit act of permission: a verbal or written agreement given in anticipation of a sexual act. While this model offers clarity and legal tractability, it reflects a relatively marginal communicative practice, most associated with specific alt-sex communities such as BDSM, and does not reflect the way consent is negotiated in the vast majority of sexual interactions (Fenner, 2017). The third, and arguably most socially prevalent, conceptualization frames consent as behavior interpreted by another person as an expression of willingness, or its absence: actions, reactions or silences that a partner interprets as indicating agreement. Under this model, the validity of consent depends entirely on the partner’s capacity and willingness to correctly read the signals being emitted.

In contexts of sexual violence prevention and judicial assessment, this third conceptualization is the most operationally relevant. Direct access to another person’s internal state is never possible; what can be investigated is which behavioral signals were emitted, how they were interpreted, and whether contextual factors or prior experiences may have shaped the behavior of the person involved (Muehlenhard et al., 2016). Coercion, within this framework, can be understood as the disregard or circumvention of refusal signals, or the absence of consent signals (Glowacz et al., 2018; Struckman-Johnson et al., 2003). Crucially, accounts of coercive experiences consistently describe situations in which implicit, and often explicit, refusal cues were present but not acted upon (Beres, 2010; DeGue & DiLillo, 2005; Gilles-Noguès et al., 2025; Struckman-Johnson et al., 2003). Sexual coercion in intimate relationships rarely takes the form of physical constraint. Rather, it generally operates through subtler mechanisms of psychological pressure: persistent re-initiation, emotional manipulation, or expressions of anger that erode the conditions for freely expressed refusal (Camilleri et al., 2009; Garrido-Macías et al., 2022). Understanding how individuals communicate consent, and how those signals are interpreted by their partner, is therefore essential to identifying the behavioral mechanisms that underlie sexual coercion (Jozkowski et al., 2014).

This framework raises a theoretically and clinically important question regarding a related phenomenon: consenting to engage in a sexual interaction despite a lack of desire for it, called sexual compliance (Gunst et al., 2024; Impett & Peplau, 2003). This phenomenon appears to be highly prevalent in heterosexual relationships (Kubota & Nakazawa, 2024; Nickull et al., 2025). Part of the literature frames sexual compliance as an autonomous relational act, unrelated to partner pressure or coercion, and associated with positive relational outcomes (Gunst et al., 2024; Nickull et al., 2025). However, empirical evidence tends to show that negative consequences are at least as frequent, if not more, and that the predictors of compliance generally point to individual vulnerabilities such as insecure attachment, low self-esteem, low sexual assertivity or difficulties in sexual communication (Darden et al., 2019; Gunst et al., 2024). Furthermore, associations have been found between sexual compliance and histories or ongoing dynamics of coercion within the relationship (Benoit & Ronis, 2022; Gilles-Noguès et al., 2024; Gunst et al., 2024; Katz & Tirone, 2010; Nickull et al., 2025). Some authors further argue that even in the absence of any coercive climate, sexual compliance may result from the internalization of gender norms; a process sometimes described as “social coercion” (Fenner, 2017; Garcia, 2021; Gunst et al., 2024; Khera et al., 2022). Whether or not sexual compliance is fully distinguishable from coercion, the fact that it appears predominantly associated with negative individual outcomes underscores the importance of deepening research on this topic; not only to inform prevention strategies and sex education, but also to support individuals in identifying alternative strategies for preserving or strengthening intimacy within relationships. Advancing this work carries direct implications for prevention, therapeutic and sexological practice.

More importantly, advancing this research agenda in French-speaking populations requires access to validated measurement tools. An instrument specifically designed to assess sexual compliance exists : the Sexual Compliance Questionnaire (Muehlenhard & Cook, 1988). This questionnaire defines compliance at the outset and focuses primarily on compliance-driven behaviors over the past year. The Internal and External Consent Scales (IECS; Jozkowski et al., 2014), offer a complementary and more fine-grained approach: rather than identifying which specific behaviors were compliant, they assess both internal and external levels of consent for a given sexual encounter, while also estimating the frequency with which individuals rely on various internal and external consent signals during their most recent sexual interaction. This makes the IECS a comprehensive and efficient instrument for research protocols on sexual consent more broadly, thereby optimizing participant response burden.

Developed through a mixed-methods approach, the scales are structured into two complementary subscales. The Internal Consent Scale assesses the subjective psychological experience associated with a sexual encounter, including dimensions of physical response, safety and comfort, arousal, agreement, and readiness; operationalizing the internal state of willingness that legal and preventive frameworks struggle to access. The External Consent Scale captures the observable behavioral signals through which consent is communicated to a partner, including nonverbal cues, verbal initiation, passivity, and, at the boundary of coercion, borderline pressure behaviors (Jozkowski et al., 2014). Together, these two scales make it possible to evaluate the degree to which an individual’s felt internal experience aligns with the signals they outwardly express for a given sexual encounter and to identify which communicative strategies individuals habitually rely upon in sexual interactions (Jozkowski et al., 2014; Marcantonio et al., 2018; Willis et al., 2019). This information is particularly valuable for understanding the conditions under which compliance occurs, and for designing contextually grounded prevention and intervention programs.

Despite the theoretical and clinical relevance of these constructs, research on sexual consent communication remains largely anchored in English-speaking, predominantly North American contexts. Validated French-language instruments are scarce, limiting the capacity of Francophone researchers and clinicians to investigate these questions in their own populations and cultural contexts. The present study aims to address this gap by examining the psychometric properties of French-language adaptations of the IECS in a French-speaking sample.

The Present Study

Thus, this study aimed at developing and evaluate the psychometric properties of French versions of the Internal Sexual Consent Scale and the External Sexual Consent Scale (IECS-FV). Specifically, the study seeks to: (a) examine the factorial structure, internal consistency, and gender invariance of the IECS-FV; and (b) investigate the associations between sexual consent–related behaviors, as assessed by these scales, and participants’ intimate relationship characteristics, as well as their sociodemographic variables.

2. Methods

2.1. Participants

A total of 649 participants took part in this study. All participants were aged between 18 and 25 years. The sample was predominantly female (73.34%), with male (22.80%), non-binary (3.24%), and other gender (0.61%) participants also represented. Regarding sexual orientation, the majority identified as heterosexual (65.02%), followed by bisexual (18.34%), other (12.17%), and homosexual (4.47%).

In terms of intimate relationship status, nearly half of participants reported being in a long-term monogamous relationship (47.44%), while the remainder were distributed across single and emerging relationship statuses. The sample was predominantly educated at high school or undergraduate level (75.03%), with a smaller proportion holding postgraduate qualifications (22.49%). Most participants reported a monthly income below 1000€ (77.81%). Full sociodemographic details are presented in Table 1.

2.2. Ethics and Procedures

The study protocol was approved by the University of Picardie Jules Verne Non-Interventional Clinical Research Ethics Committee (Approval No: 2023-25/2023-01-1) and conducted in accordance with the Declaration of Helsinki and its subsequent amendments (Wolrd health organization, 2024).

Prior to participation, all participants were presented with an information notice detailing the study’s objectives, procedures, and their rights, including the right to withdraw at any time without consequence. Informed consent was obtained from all participants before data collection. Responses were strictly anonymous, and all data were processed in compliance with the General Data Protection Regulation (GDPR).

Data reported in the present study were collected as part of a larger international research project on sexual consent and intimate relationship dynamics. Data were collected through an online questionnaire (Limesurvey team, 2024) distributed via social networks (linkedin, facebook, Instagram) and QR codes displayed across multiple university campuses in northern France, including faculties of medicine, sciences, humanities, and literature.

2.3. Measures

Sociodemographic variables. These include age, gender, intimate relationship status, educational level, and monthly income.

Intimate relationship variables. These include the number of sexual partners had in the previous 30 days; the number of sexual intercourses had in the previous 30 days; the level of satisfaction with the sexual intercourses in the previous 12 months, the level of satisfaction with the intimate partner relationships in the previous 12 months.

Internal sexual consent measure. The French version of the Internal Consent Scale (ICS-FV) was developed through the translation of the original English ICS. The original instrument (Jozkowski et al., 2014) comprises 25 items distributed across five dimensions reflecting distinct internal sexual consent–related communication processes: physical response (6 items), arousal (3 items), agreement/wantedness (5 items), safety/comfort (7 items), and readiness (4 items). All items are rated on a 4-point Likert scale ranging from strongly disagree to strongly agree. In the original validation study, the ICS demonstrated excellent internal consistency (α = .95), and all factor loadings were above .90.

External sexual consent measure. The French version of the External Consent Scale (ECS-FV) was developed through translation of the original English ECS and the addition of a 4-point Likert scale, ranging from strongly disagree to strongly agree, to include preferences measurement. The original instrument (Jozkowski et al., 2014) comprises 18 items organized into five dimensions reflecting distinct external sexual consent–related communication strategies: nonverbal behavior (5 items), passive behavior (4 items), communicative/initiator behavior (3 items), borderline pressure behavior (3 items), and non-response signals (3 items). In the original validation study, the ECS demonstrated good internal consistency (α = .84), and factor loadings ranged from .55 to .80.

Minor modifications were made to the original scale to enhance inclusivity and clarity, and to harmonize administration procedures. First, the term “vaginal-penile intercourse” was replaced by “penetrative sexual activity” (activité sexuelle pénétrative), and the phrase “or sexual activity” was replaced by “or other sexual activity” (ou autre activité sexuelle), in order to ensure the scale’s applicability to participants across diverse sexual orientations and relationship configurations. This need to make the measure more inclusive is also highlighted by Walsh et al. in a confirmatory analysis study (Walsh et al., 2019).

Second, the original instructions for the two subscales differed in their handling of participants with no experience of sexual intercourse: the ICS instructed participants to select “N/A,” while the ECS directed them to select the last response option. To improve consistency and reduce potential confusion, these divergent instructions were replaced by a unified filter question, presented once at the beginning of each subscale: “If you have never engaged in penetrative sexual activity (or other sexual activity), please select the first option only” (Si vous n’avez jamais eu d’activité sexuelle pénétrative [ou autre activité sexuelle], veuillez sélectionner la première option uniquement), this option being “I have never engaged in sexual activity” (Je n’ai jamais eu d’activité sexuelle). This harmonization was intended to standardize the administration of both subscales while minimizing participant burden.

The translation process followed a standard forward-backward translation procedure (Beaton et al., 2000). The scales were first translated from English into French by the first author, who holds a C2 level of English proficiency as attested by a score of 112/120 on the TOEFL iBT. The French version was then independently back-translated into English by a native Franco-British bilingual speaker with no prior knowledge of the original scales. Semantic equivalence between the original and back-translated versions was assessed item by item. Agreement was reached on 37 of the 43 items (86%), indicating a high level of correspondence. For the six remaining items, the translators reached consensus through discussion, and the final wording was agreed upon jointly.

Table 1 presents the original 25 ICS items (in English) and the 18 ECS items (in English), alongside their French translations, as well as the intimate relationship and sociodemographic variables assessed in the present study. Table S9 Presents the French version of the Internal and External Consent Scale (IECS-FV).

2.4. Data Analytics

First, descriptive statistical analyses were conducted for all study variables and measures, including range, mean (M), standard deviation (SD), skewness, kurtosis, and frequency distributions.

Subsequently, the suitability of the data for factor analytic procedures was assessed by examining assumptions of sampling adequacy and sphericity, using the Kaiser–Meyer–Olkin (KMO) measure and Bartlett’s test of sphericity, respectively.

Third, a confirmatory factor analysis (CFA) was conducted on the ICS data including all 25 items (CFA Model 1), using maximum likelihood estimation (MLE), to examine the scale’s dimensional structure and construct validity. MLE has been shown to provide robust and reliable parameter estimates even with categorical indicators, such as Likert-type items, particularly in studies with relatively large sample sizes (Asún et al., 2016; Rhemtulla et al., 2012). Model fit was evaluated using widely recommended fit indices (Goretzko et al., 2024), including the comparative fit index (CFI), Tucker–Lewis index (TLI), standardized root mean square residual (SRMR), and root mean square error of approximation (RMSEA). Following established guidelines (Goretzko et al., 2024; Hu & Bentler, 1999), CFI and TLI values ≥ .95 and ≥ .90 indicate excellent and good fit, respectively, whereas SRMR ≤ .08 and RMSEA ≤ .08 indicate acceptable model fit. In addition, measurement invariance of the factorial structure across gender (female vs. male participants) was assessed.

Fourth, because several standardized factor loadings in CFA Model 1 fell below the recommended threshold of 0.50, these estimates were used as criteria for item removal. Accordingly, six items with loadings below this threshold were excluded: items 8, 12, 22, and 24 from the Physical response factor, item 6 from the Arousal factor, and item 18 from the Readiness factor. In addition, based on a series of iterative CFA procedures, the Physical response and Arousal factors were combined into a single factor, termed Physical response/arousal, comprising items 2, 17, 1, and 3. The revised model resulting from these modifications was specified as CFA Model 2.

Fifth, a confirmatory factor analysis (CFA) was conducted on the ESCS data including all 18 items (CFA Model 3). As the primary fit indices (CFI and TLI) indicated poor model fit, standardized factor loadings were examined to guide item refinement. Based on this criterion, three items with loadings below 0.40 were removed: item 13 from the Communicative/initiator behavior factor, item 14 from the Borderline pressure behavior factor, and item 15 from the Non-response signals factor. Subsequently, a second CFA was performed on the revised ESCS item set, resulting in CFA Model 4.

Sixth, the internal consistency of the factor structures identified in the CFA models was evaluated by computing Cronbach’s alpha (α) for each scale factor, along with Cronbach’s α if item deleted and item–rest correlations. Cronbach’s alpha (coefficient α) is a reliability index assessing the internal consistency of a scale, with values ranging from 0 to 1. Conventionally, values ≥ .70 are considered acceptable for exploratory research, ≥ .80 for established research contexts, and ≥ .90 for high-stakes clinical applications (Bland & Altman, 1997; Nunnally, 1978; Zijlmans et al., 2018). In the present study, reliability was estimated for the total score of each sexual consent–related behavioral dimension. Cronbach’s α if item deleted indicates the reliability coefficient that would be obtained if a given item were removed from the scale, thereby informing item contribution to overall consistency. Item–rest correlation reflects the association between each item and the total score computed from the remaining items, with values ≥ .20 generally considered acceptable (Bland & Altman, 1997; Nunnally, 1978; Zijlmans et al., 2018).

Seventh, to examine the relationships between sexual consent–related behaviors (as derived from CFA Model 2 and CFA Model 4) and participants’ intimate relationship measures and sociodemographic variables, we conducted analyses of variance (ANOVA) for nominal variables (i.e., gender and sexual orientation) and Pearson correlation analyses for the remaining ordinal variables.

All analyses were performed using Jamovi statistical software (version 2.6; The Jamovi Project, 2025) for descriptive and correlational analyses, and the R statistical environment, specifically the lavaan package (version 0.6-19; Rosseel, 2012), for confirmatory factor analyses.

3. Results

3.1. Descriptive Statistics

Table 1 presents a summary of the descriptive statistics for participants’ responses to the initial 25 items of the ICS-FV, the 18 items of the ECS-FV, the four intimate relationship measures, and the six sociodemographic variables.

3.2. Confirmatory Factorial Analysis (CFA) Results

3.2.1. CFA Related to the ICS

The assumptions of sampling adequacy and sphericity were met (KMO = 0.87; Bartlett’s test of sphericity: χ² [df = 153] = 3200, p < .001) (Cerny & Kaiser, 1977).

The CFA Model 1 (ICS data comprising 25 items and five factors) showed acceptable fit indices: χ² = 844, df = 220, p < .001; CFI = .93; TLI = .92; SRMR = .045; RMSEA = 0.066 (95% CI [.061–.070]). However, six standardized factor loadings—corresponding to items 8, 12, 22, 24, 6, and 18—were below the 0.50 threshold.

The overall CFA Model 2 (ICS data including 19 items and four factors) demonstrated excellent fit indices, and an improvement over CFA Model 1: χ² = 536, df = 129, p < .001; CFI = .95; TLI = .94; SRMR = .036; RMSEA = 0.069 (95% CI [.063–0.075]). Figure 1 presents the CFA path diagram for CFA Model 2. Standardized factor loadings of the observed variables on the latent constructs (see Figure 1) ranged from 0.62 to 0.90 and were all statistically significant (p < .001), indicating that each indicator reliably reflects its corresponding latent construct (Goretzko et al., 2024; Hu & Bentler, 1999). The variances of the observed variables (see Table S2 in the Supplementary Material) were also statistically significant (p < .001), suggesting adequate variability across measures (Goretzko et al., 2024; Hu & Bentler, 1999).

Nunnally, J. C. (1978). Cronbach’s Alpha. Journal of Consumer Research, 16, 297-334.

Consequently, the final ICS-FV is based on the factorial structure derived from CFA Model 2, comprising 19 items across four dimensions. Detailed statistics for CFA Model 2, including factor loadings and factor covariances, are presented in Tables S1 and S2 of the Supplementary Material.

3.2.2. CFA Related to the ECS

The assumptions of sampling adequacy and sphericity were met (KMO = 0.96; Bartlett’s test of sphericity: χ² [df = 300] = 10,093, p < .001) (Cerny & Kaiser, 1977).

The overall CFA Model 3 (ESCS data including 18 items) showed poor fit indices: χ² = 447, df = 93, p < .001; CFI = .86; TLI = .82; SRMR = .063; RMSEA = 0.076 (95% CI [.069–.083]).

The overall CFA Model 4 (ESCS data including 15 items) demonstrated an acceptable fit to the observed data and an improvement over CFA Model 3: χ² = 281, df = 80, p < .001; CFI = .92; TLI = .90; SRMR = .049; RMSEA = 0.062 (95% CI [.054–0.070]). Figure 2 presents the CFA path diagram for CFA Model 4. Standardized factor loadings of the observed variables on the latent constructs (see Figure 2) ranged from 0.47 to 0.91 and were all statistically significant (p < .001), indicating that each indicator adequately reflects the underlying latent construct (Goretzko et al., 2024; Hu & Bentler, 1999). Notably, only two items (items 9 and 5) presented loadings below 0.50. The variances of the observed variables (see Table S6 in the Supplementary Material) were also statistically significant (p < .001), suggesting sufficient variability across the measures (Goretzko et al., 2024; Hu & Bentler, 1999).

Consequently, the final ECS-FV is based on the factorial structure derived from CFA Model 4, comprising 15 items across five dimensions. Detailed statistics for CFA Model 4, including factor loadings and factor covariances, are presented in Tables S5 and S6 of the Supplementary Material.

3.3. Gender (Female vs. Male) Invariance Tests

For the ICS-FV, the configural invariance model demonstrated an acceptable fit (CFI = .93; TLI = .92; SRMR = 0.050; RMSEA = 0.079). However, the metric invariance test indicated that the factor loadings were not equivalent across genders, suggesting statistically significant differences in the meaning of the three modeled constructs between female and male participants (Δχ² = 878.95, df = 292, p < .001).

For the ECS-FV, the configural invariance model also showed an acceptable fit (CFI = .91; TLI = .90; SRMR = 0.060; RMSEA = 0.066). In contrast to the ICS-FV, the metric invariance test indicated that factor loadings were invariant across genders, suggesting that the meaning of the three modeled constructs did not differ statistically between female and male participants (Δχ² = 387.78, df = 160, p = .070).

3.4. Internal Reliability

As reported in Table 2, the internal reliability indices for the final ICS-FV (CFA Model 2; 19 items and four factors) and the final ECS-FV (15 items and five factors) indicated good reliability. Specifically, Cronbach’s alpha values were ≥ .70 across all factors, Cronbach’s α if item deleted supported the retention of the observed items, and item–rest correlations were ≥ .30, further confirming satisfactory internal consistency.

3.5. Comparisons Between the ICS-FV and the ECS-FV Scores

There was a statistically significant difference in the mean scores across the four internal sexual consent (ICS) behavioral dimensions, χ² (df = 3) = 281, p < .001. According to participants’ ratings, the most frequently endorsed ISC behavior was Agreement/wantedness (M = 3.71), followed by Safety/comfort (M = 3.53), Physical response/arousal (M = 3.48), and Readiness (M = 3.44), which was the least endorsed. Further details regarding this finding are provided in Table S4 of the Supplementary Material.

There was also a statistically significant difference in the mean scores across the five external sexual consent (ECS) behavioral dimensions, χ² (df = 3) = 936, p < .001. According to participants’ ratings, the most frequently endorsed ECS behavior was Non-verbal behavior (M = 3.43), followed by Passive behavior (M = 3.37), Communicative/initiator behavior (M = 3.26), Borderline pressure (M = 2.72), and Non-response signals (M = 2.03), which was the least endorsed. Further details regarding this finding are provided in Table S8 of the Supplementary Material.

3.6. Correlations Between the Intra Scales’ Specific Sexual Consent Behaviors

Table 3 presents the correlations between the ICS-FV and ECS-FV sexual consent behaviors, as well as the associations between these behaviors and the other modeled study variables.

All correlations among the four ICS-FV sexual consent behaviors were positive and statistically significant. Similarly, correlations among the ECS-FV sexual consent behaviors were all statistically significant; most were positive, with the exception of Non-response signals, which showed a positive association with Passive behavior but negative associations with all other ECS-FV sexual consent behaviors.

3.7. Correlations Between the Inter Scales’ Specific Sexual Consent Behaviors

As shown in Table 3, all ICS sexual consent behaviors (Physical response/arousal, Agreement/wantedness, Safety/comfort, and Readiness) are significantly correlated with the ECS sexual consent behaviors (Non-verbal behavior, Passive behavior, Communicative/initiator behavior, Borderline pressure behavior, and Non-response signals). Most correlations were positive. The only exception was Non-response signals, which, although statistically significant, was negatively associated with the other sexual consent behaviors.

3.8. Relationships Between the ICS-FV Sexual Consent Behaviors, the ECS-FV Sexual Consent Behaviors, the Participants Gender and Sexual Orientation

ANOVA and follow-up t-test analyses (see Tables S4 and S8 in the Supplementary Material) were conducted to examine whether participants’ gender and sexual orientation were associated with their scores on ICS and ECS sexual consent behaviors. Overall, few differences emerged according to gender and sexual orientation in the endorsement of ICS behaviors. Significant differences were observed only for Physical response/arousal and Agreement/wantedness. Regarding Physical response/arousal, male participants reported significantly higher scores than female participants (M = 3.49 [0.02] vs. M = 3.42 [0.02], t(645) = 4.45, p = .001, d = 0.41), indicating a greater likelihood of endorsing this behavioral strategy in communicating sexual consent. Similarly, for Agreement/wantedness, male participants also scored significantly higher than female participants (M = 3.81 [0.03] vs. M = 3.68 [0.02], t(645) = 3.02, p < .014, d = 0.28), suggesting a higher endorsement of this strategy among male participants.

With regard to ECS behaviors, few gender and sexual orientation differences were also observed. A significant difference emerged only for Communicative/initiator behavior, with male participants reporting higher scores than female participants (M = 3.48 [0.06] vs. M = 3.18 [0.03], t(645) = 4.145, p < .001, d = 0.39). This indicates that male participants were more likely than female participants to report using this behavioral strategy to communicate sexual consent.

3.9. Correlation Between the ICS-FV Sexual Consent Behaviors, the ECS-FV Sexual Consent Behaviors, the Participants’ Relationship, and Remaining Sociodemographic Variables

As shown in Table 3, ICS Physical response/arousal was significantly and positively correlated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, and the level of intimate relationship commitment. ICS Agreement/wantedness was significantly and positively correlated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, and the level of intimate relationship commitment. ICS Safety/comfort was significantly and positively correlated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, the level of intimate relationship commitment, and income. ICS Readiness was significantly and positively correlated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, and the level of intimate relationship commitment. ECS Non-verbal behavior was significantly and positively correlated to sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, and the level of intimate relationship commitment. ECS Passive behavior was significantly and positively correlated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction; it was significantly and negatively correlated to age, educational level, and income. ECS Communicative/initiator behavior was significantly and positively correlated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, and the level of intimate relationship commitment. ECS Borderline pressure was significantly and positively correlated with sexual satisfaction and romantic relationship satisfaction; was significantly and negatively correlated with age and educational level. ECS Non-response signals was significantly and negatively correlated with sexual satisfaction, romantic relationship satisfaction, and with the level of intimate relationship commitment.

As shown in Table 3, ICS Physical response/arousal was significantly and positively correlated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, and level of intimate relationship commitment. Similarly, ICS Agreement/wantedness was significantly and positively associated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, and intimate relationship commitment. ICS Safety/comfort was also significantly and positively correlated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, intimate relationship commitment, and income. ICS Readiness showed significant and positive associations with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, and intimate relationship commitment.

Regarding ECS behaviors, Non-verbal behavior was significantly and positively correlated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, and intimate relationship commitment. Passive behavior was significantly and positively associated with sexual intercourse frequency, sexual satisfaction, and romantic relationship satisfaction, while showing significant negative correlations with age, educational level, and income. Communicative/initiator behavior was significantly and positively correlated with sexual intercourse frequency, sexual satisfaction, romantic relationship satisfaction, and intimate relationship commitment. Borderline pressure was significantly and positively associated with sexual satisfaction and romantic relationship satisfaction, and significantly and negatively correlated with age and educational level. Finally, Non-response signals were significantly and negatively correlated with sexual satisfaction, romantic relationship satisfaction, and intimate relationship commitment.

4. Discussion

4.1. Overview of the Study Objectives and Principal Findings

The present study aimed to develop and validate French versions of the Internal Consent Scale and External Consent Scale (IECS-FV) among French-speaking emerging adults, while also examining their factorial validity, reliability, gender invariance, and associations with sociodemographic and intimate relationship variables. Overall, findings provide substantial support for the psychometric adequacy of both French adaptations, while also highlighting important cultural and gender-related nuances in the measurement and expression of sexual consent. These findings contribute significantly to the Francophone literature by offering validated instruments capable of assessing internal and external dimensions of sexual consent communication in young adults.

Consistent with prior conceptualizations of sexual consent as both an internal psychological state and an externally communicated interpersonal process (Jozkowski et al., 2014; Muehlenhard et al., 2016), the present results suggest that sexual consent among French-speaking young adults can be reliably measured through multidimensional constructs. However, modifications to the original factorial structures were required, particularly for the ICS-FV, indicating that some aspects of sexual consent may be culturally or developmentally specific.

4.2. Psychometric Properties of the ICS-FV

The French adaptation of the Internal Consent Scale (ICS-FV) demonstrated strong psychometric performance after refinement. While the original five-factor structure proposed by Jozkowski et al. (2014) initially showed acceptable fit, six items displayed insufficient loadings, necessitating item removal and restructuring. The resulting 19-item, four-factor model demonstrated excellent fit and robust reliability.

Importantly, the merger of Physical Response and Arousal into a single factor suggests that in this French-speaking sample, physiological sexual responses and subjective arousal may be experienced less distinctly than in the original North American validation sample. This may reflect developmental characteristics of emerging adulthood, where bodily and emotional experiences of sexuality are highly intertwined (Quinn-Nilas et al., 2016), or sociocultural differences in how sexual readiness and arousal are conceptualized (Willis et al., 2019).

The retained dimensions (Physical Response/Arousal, Agreement/Wantedness, Safety/Comfort, and Readiness) align closely with theoretical models positioning consent as a dynamic internal process involving desire, autonomy, and emotional security (Marcantonio et al., 2018). Particularly noteworthy is the prominence of Agreement/Wantedness and Safety/Comfort, which emerged as the highest-rated dimensions, reinforcing the centrality of subjective willingness and psychological safety in consensual sexual experiences (Muehlenhard et al., 2016).

The strong internal consistency across all factors further supports the ICS-FV as a reliable measure for assessing nuanced internal sexual consent experiences in French-speaking populations.

4.3. Psychometric Properties of the ECS-FV

The ECS-FV also demonstrated satisfactory psychometric properties, though refinement was necessary. The revised 15-item model showed acceptable to good fit, supporting the multidimensionality of external sexual consent communication.

The preservation of five distinct factors (Non-verbal behavior, Passive behavior, Communicative/Initiator behavior, Borderline pressure, and Non-response signals) largely replicates the original structure (Jozkowski et al., 2014), suggesting stronger cross-cultural stability in external consent communication strategies compared to internal consent experiences.

Non-verbal behavior emerged as the most commonly endorsed external strategy, consistent with prior literature indicating that sexual consent is often communicated behaviorally rather than verbally in heterosexual and emerging adult relationships (Edwards et al., 2022; Willis et al., 2019).

Passive behavior’s relatively high endorsement may also reflect the prevalence of indirect consent strategies, which may overlap with sexual compliance dynamics (Darden et al., 2019; Muehlenhard et al., 2016). Meanwhile, lower endorsement of Non-response signals suggests that overt ambiguity or silence may be less normative but remains clinically relevant given its negative associations with relational satisfaction (Jozkowski et al., 2014).

Overall, the ECS-FV appears well-suited for assessing external sexual consent communication patterns in French-speaking youth.

4.4. Gender Invariance and Gender Differences

Gender invariance analyses yielded contrasting findings for the two scales. The ECS-FV demonstrated metric invariance across male and female participants, indicating that external sexual consent communication behaviors are expressed similarly across genders. This suggests that communication strategies may be relatively stable across gender groups, in accordance with what is reported in the literature (Muehlenhard et al., 2016).

In contrast, the ICS-FV failed to demonstrate metric invariance, indicating that the internal psychological experience of sexual consent may differ meaningfully between men and women. This finding is theoretically coherent with gendered sexual socialization literature: women’s consent experiences are likely shaped by a convergence of factors, including the risk of social stigmatization, the possibility of physical pain, and internalized norms of “sexual debt”, that introduce layers of complexity less salient in men’s experiences (Beres, 2010; Conroy et al., 2015; Geary et al., 2013). Notably, this framing does not imply that men’s consent experiences are uniformly straightforward; research on sexual compliance suggests that a meaningful proportion of men also engage in unwanted sexual activity, often linked to adherence to stereotypical gender roles (Peterson, 2024). Nevertheless, the relational, physical, and normative asymmetries women face in sexual negotiation may generate a qualitatively distinct phenomenology of consent (Fenner, 2017).

Observed gender differences further supported this interpretation. Male participants reported higher levels of Physical Response/Arousal, Agreement/Wantedness, and Communicative/Initiator behavior. These findings align with traditional gender norms that position men as sexual initiators and women as gatekeepers (Garcia, 2021), though effect sizes were modest.

These results emphasize the importance of considering gender-specific dynamics in both sexual consent research and prevention programming.

4.5. Relationships Between Internal and External Consent Dimensions

Significant positive correlations between all internal consent dimensions and most external consent behaviors support the theoretical coherence of the IECS framework. Individuals reporting stronger internal willingness, readiness, and safety were generally more likely to engage in active external consent communication.

This coherence between internal and external consent is particularly important, as discrepancies between felt consent and expressed consent may represent conditions under which sexual compliance or coercion are more likely to occur (Gilles-Noguès et al., 2024; Hust et al., 2017; Katz & Tirone, 2010).

Notably, Non-response signals were negatively associated with all positive consent dimensions, suggesting that silence or ambiguity may function as markers of lower consent congruence and potentially greater vulnerability.

These findings support the utility of simultaneously measuring both internal and external consent processes to better understand sexual agency and relational dynamics.

4.6. Associations with Sociodemographic and Intimate Relationship Variables

A consistent pattern emerged linking more consistent consent profiles with more positive intimate relationship functioning. Internal consent dimensions were positively associated with sexual frequency, sexual satisfaction, romantic satisfaction, and relationship commitment. Similarly, active external consent strategies (e.g., non-verbal and communicative behaviors) were positively related to these variables. These findings align with prior evidence suggesting that consensual, communicative sexual interactions contribute to sexual satisfaction (Marcantonio et al., 2020).

Conversely, passive behaviors and non-response signals were associated with lower socioeconomic indicators, younger age, and lower relationship quality. This may indicate that less empowered consent communication is linked to developmental immaturity, reduced relational resources, or broader social inequalities (Sánchez-Ocaña et al., 2026).

Interestingly, Borderline pressure behaviors were positively associated with sexual and relational satisfaction, a finding that warrants caution. It is possible that some borderline pressure items may capture playful persuasion rather than coercion in certain relational contexts. However, given the literature on normalization od coercive behaviors (Burt, 1980; Camilleri et al., 2009; Wilson, 2023), this dimension requires further conceptual clarification.

4.7. Limitations

Several limitations should be acknowledged. First, the sample was predominantly female, heterosexual, and mainly university-affiliated, which may limit generalizability to broader French-speaking populations. Second, the cross-sectional design precludes causal inference regarding the relationships between sexual consent behaviors and relational outcomes. Third, all data were self-reported, raising potential concerns regarding recall bias and social desirability effects (King, 2024). Fourth, the exclusion of non-binary participants from invariance analyses due to sample size limitations restricts understanding of sexual consent measurement across more diverse gender identities. Finally, while translation procedures were rigorous, some structural modifications suggest possible cultural adaptation effects that warrant replication.

5. Future Research Directions

Future studies should examine the IECS-FV in more diverse populations, including older adults, LGBTQ+ communities, and clinical populations. Longitudinal designs are needed to assess temporal stability and predictive validity. Further investigation into internal-external consent discrepancies may improve understanding of sexual compliance and coercion. Additionally, qualitative research could help clarify culturally specific meanings of passive, borderline, and non-response consent behaviors within Francophone contexts. Cross-cultural comparisons between French-speaking and other populations may also advance global sexual consent theories.

6. Conclusion

The present study provides evidence for the validity and reliability of French versions of the Internal and External Sexual Consent Scales among emerging adults. These instruments represent valuable tools for advancing Francophone research, prevention, and clinical practice related to sexual consent, sexual compliance, and intimate relationship functioning.

By capturing both internal experiences and external communication strategies, the IECS-FV offers a comprehensive framework for understanding sexual consent as a multidimensional psychosocial process. These findings support the broader conceptualization of sexual consent as central not only to sexual violence prevention, but also to relational health, autonomy, and sexual well-being.