Sexting Motivation Scale (EMS) in Peruvian Youth

Abstract

The sharing of explicit sexual images through virtual platforms has grown exponentially in recent generations, causing various issues such as low self-esteem, sextortion, and cyberbullying, among others. Therefore, it is imperative to have a deeper understanding of this issue. This research aims to construct an instrument that allows for the identification of the motivations that lead to the practice of sexting in young university students. The research had two phases: conducting an Exploratory Factor Analysis, which included 320 university students (48.2% female and 51.8% male) aged between 18 and 30 (M = 20.4; SD = 2.62). In the second phase, 1056 university students (55.9% female and 44.1% male) within the same age range as the first phase participated (M = 22.38; SD = 2.64), for whom a Confirmatory Factor Analysis (CFA) was conducted. The final version of the scale consisted of a unidimensional model comprising eight items that assess internal and external motivations leading to the practice of sexting, with an adequate fit index (CFI = 0.991, TLI = 0.988, and RMSEA = 0.068). Additionally, an omega coefficient of 0.93 was found, indicating adequate reliability. It was concluded that the Sexting Motivation Scale (EMS by its initials in Spanish) demonstrates good reliability and construct validity, making it suitable for measuring motivation for sexting.

1. Introduction

Social media has significantly impacted adolescents and young adults, transforming their interactions in recent years. Even during the COVID-19 pandemic, the use of social media increased significantly, becoming an advantage as it allowed for virtual connection during social distancing [1] (Venegas-Vera et al., 2020). However, it also poses a potential threat, bringing about new attitudes and behaviors in current generations, such as low mood [2,3] (Berry et al., 2018; Szlyk et al., 2023), cyber addiction, cyberbullying, and grooming [4,5,6,7] (Gupta, Kattapuram, and Patel 2020; Siegmund 2020; Kross et al. 2021; Ruiz et al. 2021) Additionally, there is sexting, which is defined as the act of sharing, through electronic devices, messages that contain photos or videos and show explicit sexual content [8,9] (Ruiz et al., 2021; Van Ouytsel et al., 2019).

According to [10] Mori et al. (2020), 38.3% of adolescents and young adults in Canada between the ages of 14 and 20 engage in sexting. Similarly, in Spain, it was found that more than a third of adolescent girls aged 14 to 18 practice sexting [11,12,13] (Gámez-Guadix et al., 2017; Ojeda et al., 2020; Villacampa, 2017). In a recent meta-analysis examining studies from the United States, Europe, Australia, Canada, South Africa, and South Korea, it was found that the prevalence of sending sexual content was 14.8%, receiving was 27.4%, forwarding sexual content without consent was 12.0%, and receiving forwarded sexual messages was 8.4% [14] (Madigan et al., 2018a).

In Peru, the practice of sexting is not regulated, but the sending of sexual material without the consent of the author is a crime that can lead to imprisonment for up to 5 years, with the consensual mode being more common in men, while the forced mode is more common in women [15,16] (Gómez-Galindo et al., 2022; Vega-Gonzales et al., 2021). According to statistical information, it was observed that 46.2% accessed inappropriate content on their computers, 24.9% practiced sexting, and 12.3% had suffered sextortion, and 95.5% of university students stated that they had participated in sexting at some point, thus showing that the prevalence of this problem is evident [17] (Giménez-Gualdo et al., 2022).

Among all age groups, adolescents and young adults are particularly vulnerable to sexting as they are exposed to virtualization and digital connectivity [18] (Bianchi et al., 2021), making it a platform to explore their sexuality [12] (Ojeda et al., 2020). This stage marks the beginning of sexual development, and individuals start experimenting with dating and forming romantic relationships [19] (Sawyer et al., 2018). Although many studies suggest that sexting can be practiced safely if strategies are used to reduce negative consequences [20,21] (Molla Esparza et al., 2020; Strasburger et al., 2019), it can still lead to social, physical, psychological, and often legal consequences [12] (Ojeda et al., 2020). Women are the most vulnerable and can be negatively affected [22] (Arias, 2018). Additionally, these practices expose individuals to various risks [23] (Buzi, 2019), such as cyberbullying and grooming, among others, with the most common being sexting through non-consensual third-party distribution, prevalent among 8.4% to 15.6% of youth [20] (Molla Esparza et al., 2020). Likewise, it can be used for revenge by an ex-partner to coerce the victim into resuming the relationship or to extort victims for money by threatening to distribute the images if demands are not met [18,24] (Bianchi et al., 2021; Ross et al., 2019). This leads to negative psychological consequences [25,26] (Gassó et al., 2019; Sciacca et al., 2023), starting with the fear of unauthorized dissemination of body images [25,27] (Garrido-Macías et al., 2023; Klettke et al., 2019). Similarly, [28] Milton et al. (2019) note that individuals may experience cyber victimization, depression, anxiety, and low self-esteem.

In this context, it is essential to have instruments that allow for a deep analysis of this problem. Motivation is a key factor in decision making and subsequent actions, as pointed out by the Self-Determination Theory, which distinguishes between intrinsic and extrinsic motivations to explain human behavior [29] (Deci, E. L., & Ryan, 2013). In the case of sexting, motivations may be driven as much by the desire for self-affirmation and personal exploration as by social pressure or external expectations. In addition, Social Learning Theory suggests that behaviors are learned through observation and imitation of others, which contributes to the normalization of sexting in certain environments (Bandura, 1986). In Peru, studies on this topic are limited, as there is only one validated instrument, the Sexting Behavior Scale [30] (Chacón-López et al., 2016), which was adapted to the Peruvian population by [31] Rios-Alvites et al., (2022) in a sample of university students from Lambayeque. This evidences the need to develop and validate a specific tool to measure the motivation behind sexting in the local context. This evidences the need to develop and validate a specific tool that measures the motivation behind sexting in the local context, as pointed out by [20] Esparza et al. (2020), who highlight the urgency of investigating the factors that drive this practice in Peruvian youth. In light of this, the need to understand the motivation for the practice of sexting emerges.

Conceptual Delimitation of Motivation for Sexting

The act of sharing (sending and receiving) messages, photographs, or videos depicting the body in explicitly sexual situations is often associated with immaturity and emotional instability [25,32,33,34,35] (Barroso et al., 2021; Chacón-López et al., 2019; Choi et al., 2019; Gassó et al., 2019; Rodríguez-Castro et al., 2018). In the context of this study, the relational needs that tend to be satisfied through social media are primarily related to emotional connection, social validation, and belonging to a group. Young people, especially teenagers and young adults, use social media to maintain and strengthen interpersonal relationships, as well as to establish emotional and sexual bonds [36] (Van Ouytsel et al., 2020). These platforms allow for constant interaction and message exchange, which satisfies the need for social acceptance and identity affirmation. Regarding the most used platforms, it is observed that highly visual social media, such as Instagram and Snapchat, are more frequent in this population compared to text-based platforms like Facebook. These networks, being more visual, facilitate the expression of body image and the exploration of sexuality, which can contribute to internal and external motivations associated with sexting [14,25] (Gassó et al., 2019; Madigan et al., 2018b).

In this context, motivations for sexting can be classified into internal and external, as suggested by previous studies [37,38] (Amorós 2014; Bragard and Fisher 2022). From a psychological perspective, intrinsic motivation is associated with the desire for personal exploration, sexual experimentation and identity reaffirmation [39,40] (Baumgartner et al., 2014; Currin, 2022), while extrinsic motivation is a product of social influences, such as peer pressure or the need for acceptance or substance use [41] (Houck et al., 2014); on the other hand, external motivations are situations that occur in the social environment, such as peer pressure, the influence of problematic friends, and substance use, among others [42,43,44] (Chan & Wu-Ouyang, 2023; Crimmins & Seigfried-Spellar, 2014; Temple et al., 2013). These motivations can lead to concrete actions, such as sending intimate content, the consequences of which will depend on the level of control the individual has over the decision. Through Reinforcement Theory [45] (Skinner, 1965), it is explained that sexting can be perpetuated due to reinforcing responses, such as social approval or a momentary increase in self-esteem. Therefore, understanding these factors is crucial to design effective prevention and awareness strategies, especially in adolescents and young adults.

2. Methodology

2.1. Methodological Design

This research is non-experimental since no variable manipulation occurred, quantitative because numerical data were collected, cross-sectional because it was conducted at a single point in time, and instrumental in design because the psychometric properties of an instrument will be examined [46] (Ato et al., 2013).

2.2. Participants

For the Exploratory Factor Analysis (EFA), 320 university students participated (53.0% from private universities and 47.0% from national universities) from the San Martín region of Peru, of which 48.2% were female and 51.8% were male, aged between 18 and 30 years (M = 20.4, SD = 2.62). The minimum age of the participants was 18 years old, so they were considered adolescents according to the WHO and the Peruvian Technical Health Standard [47] (MINSA, 2019).

Regarding the Confirmatory Factor Analysis (CFA), 1056 young university students and adolescents participated (57.2% from private universities and 42.8% from national universities; 55.9% female and 44.1% male; 33.6% from the Coast, 33.0% from the Highlands, and 33.4% from the Jungle of Peru), who regularly attended classes, were within an age range of 18–30 years (M = 22.38, SD = 2.64), and could access the research, and they were selected through non-probabilistic snowball sampling [48] (Atkinson, 2001).

2.3. Instruments

Sexting Motivation Scale (EMS)

The Sexting Motivation Scale (EMS) was specifically designed for this study to measure internal and external motivations related to sexting behavior. The EMS consists of 8 items that assess various motivations, such as partner pressure, the need to impress, and self-affirmation, among others. Participants responded using a Likert scale ranging from ‘Never’ (1) to ‘Many times’ (4).

2.4. Procedure and Ethical Considerations

The research began with a bibliographic review of various studies on different platforms, such as Scielo, Google Scholar, Redalyc, Sciencedirect, Alicia, Ebsco, and Scopus, to estimate theoretical support regarding sexting in adolescents and young adults. Subsequently, the analysis involved five expert judges: clinical psychologists with over 5 years of experience and familiarity with the subject.

Permission was sought from universities to enter their premises and recruit the sample. For this purpose, a virtual questionnaire was developed (using the Google Forms tool) consisting of 3 sections:

• The first section explained the purpose of the research and presented the informed consent form, giving the participants the freedom to choose whether to participate. If the participant chose not to participate, the questionnaire was automated to skip all questions.
• The second part asked participants to provide information about sociodemographic data, with different mandatory questions for the research. Participants were not required to reveal their identities to guarantee anonymity.
• The third and final section detailed the questions with the appropriate response options.

The sample was collected by conducting face-to-face sessions, class by class, and through referrals from students who recommended their friends to participate voluntarily, respecting their privacy and confidentiality as established in the Helsinki Treaty [49] (Association, 2013). In addition, prior to the application of the survey, this study was approved on the record N° 2023-CE-EPG-00132 dated 9 March by the ethics committee of the Graduate School of the Universidad Peruana Unión.

In both instances, students from public and private universities in the San Martín region of Peru who enrolled between the first and fifth academic year were considered. Although the sample is smaller than 400, it is important to keep in mind that if the correlations are strong and the number of factors is small, the sample size can be smaller [50] (Tabachnick, 2021). Likewise, if the communalities are high and if there are no cross-loadings, the sample can also be small [51,52] (Costello & Osborne, 2005; Thompson, 2004), also considering the principle that the sample quantity is derived from the nature of the data. Thus, 331 data were retained, which is an adequate number to perform the AFE, since it reports high communalities.

2.5. Data Analysis

After obtaining the data from Google Forms, a database was created in Microsoft Excel, and it was recoded before being transferred to the statistical software SPSS version 26 and RStudio (version 2024.12.1+563).

The study began with a descriptive analysis of the items, for which the mean, standard deviation, kurtosis, skewness, and corrected homogeneity index were calculated. This initial analysis was conducted using SPSS.

Subsequently, the data were analyzed using RStudio. For the Exploratory Factor Analysis (EFA), polychoric correlations, unweighted least squares, and Promax rotation were used because the dimensions were correlated. For the Confirmatory Factor Analysis (CFA), the Comparative Fit Index (CFI), Tucker–Lewis Index (TLI), and Root Mean Square Error of Approximation (RMSEA) were used. The minimum required values for accepting the model were CFI and TLI ≥ 0.90 [53,54] (Bentler & Bonett, 1980; Tucker & Lewis, 1973) and RMSEA < 0.08 [55] (MacCallum et al., 1996). For reliability analysis, McDonald’s total omega coefficient was calculated with a minimum accepted value of 0.65 [56] (Katz, 2007).

3. Results

3.1. Content Validity

Five clinical psychologists with experience in the topic and familiarity with the study population were consulted for content validity. They were presented with the scale containing all items, and based on their analysis, modifications were made to the recommended items. Regarding the results of Aiken’s V, the minimum score was 0.8, and the maximum was 1, which is considered acceptable according to [57] Merino-Soto (2018).

3.2. Descriptive Analysis of Items

The mean, which had a score between 2 and 3, was determined to measure the difficulty of the indicators. Additionally, skewness and kurtosis values were examined to observe if the items have an adequate distribution, ranging between 0.00 and 0.54 for skewness and −1.35 and −0.82 for kurtosis, respectively. These values are within ±2, indicating that the items are within the appropriate parameters [58,59] (Gil-Garcia et al., 2018; Hair et al., 2014). Furthermore, the corrected homogeneity indices (CHIs) ranged from 0.52 to 0.73, meeting the criterion [60] Stephan et al. (1941), who suggest that the appropriate value should be >0.20 (see

Table 1. Descriptive analysis of items.

No. Items.	Mean	SD	Skewness	Kurtosis	CHI
1	2.29	1.06	0.16	−1.22	0.68
2	2.03	1.00	0.51	−0.95	0.55
3	2.02	0.95	0.48	−0.83	0.64
4	2.40	0.99	0.00	−0.83	0.64
5	2.21	1.07	0.34	−1.15	0.70
6	2.33	1.10	0.14	−1.16	0.70
7	2.18	0.98	0.31	−0.98	0.60
8	2.09	1.01	0.40	−1.06	0.52
9	1.97	0.97	0.52	−0.92	0.57
10	2.13	1.01	0.32	−1.10	0.67
11	2.18	1.02	0.33	−1.06	0.67
12	2.08	0.95	0.33	−0.99	0.69
13	2.15	1.02	0.28	−1.17	0.70
14	2.28	1.04	0.15	−1.22	0.68
15	2.29	1.05	0.09	−1.28	0.66
16	3.00	1.11	0.02	−1.35	0.73

).

3.3. Exploratory Factor Analysis

An Exploratory Factor Analysis (EFA) was conducted to assess the adequacy of the Sexting Motivation Scale (SMS). The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy was found to be 0.97, and a χ² (120) = 7598.332, p < 0.000, was obtained in Bartlett’s test of sphericity [61,62] (Bartlett, 1950, 1951), indicating that the sample was suitable for EFA [52,63] (Kaiser 1970; Thompson 2004b). This suggests that the original variables, from the Sexting Motivation Scale, could be factorized. Additionally, parallel analysis revealed a bidimensional structure, indicating the presence of both internal and external motivators. Furthermore, the psychometric quality of the items was analyzed, revealing that three of them were not acceptable according to the theory of double factorial weight [64] (Reise et al., 2010). Specifically, items 1, 14, and 16 exhibited higher factorial loadings on the second dimension, leading to their exclusion from the analysis.

Table 2. Descriptive analysis of items.

	Internal Motivators	External Motivators	H2
P1	0.77		0.86
P2	0.79		0.83
P3	0.77		0.82
P4	0.69		0.78
P5	0.69		0.82
P6	0.69		0.84
P7	0.76		0.82
P8	0.75		0.86
P9	0.67		0.80
P10	0.71		0.83
P11	0.66		0.83
P12	0.70		0.85
P15		0.54	0.67
P17		0.75	0.79
P18		0.87	0.86
P19		0.73	0.85

displays the items considered for the final version of the scale, which demonstrate adequate values falling within the established criteria.

3.4. Confirmatory Factor Analysis of Invariance

Initially, a two-factor model with 16 items was created (Figure 1), where optimal results were examined for both the CFI and TLI, exceeding 0.95; however, the RMSEA value for this model was 0.106, which is higher than expected. After obtaining the results of model one, a new model was proposed, in which, after statistical analysis, eight items were eliminated. Items P-4, P-6, and P-7 were removed due to their theoretical similarity. Additionally, items P-8 and P-11 were located in the internal motivation dimension, but the other dimension (external motivations) was theoretically described. Regarding items P-13, P-14, and P-15, they were eliminated because they had a factorial loading lower than 0.3 [65] (Ferrando & Anguiano-Carrasco, 2010), which did not allow us to obtain an RMSEA lower than 0.80, turning the EMS into a one-dimensional instrument (Figure 2). Thus, the following results were obtained: CFI = 0.991, TLI = 0.988, and RMSEA = 0.068. These fall within the appropriate parameter ranges without the need to perform error covariances (

Table 3. Fit indices in confirmatory factor analysis according to Models 1 and 2 of the Sexting Motivation Scale (EMS).

	(p)X2	Df	x2/df	CFI	TLI	RMSEA	SRMR
Model 1 (16 items)	0.000	120	33,323.862	0.962	0.962	0.106	0.054
Model 2 (8 items)	0.000	28	11,699.186	0.991	0.988	0.068	0.027

). It should be noted that, despite being unidimensional, the instrument includes indicators that measure internal and external factors. The instrument includes items that measure the practice of sexting as a result of personal motivation (items 1, 3, 5, and 6) and behaviors resulting from an external conditioning factor, as is practiced due to social pressure (items 2, 4, 7, and 8) (Appendix A).

3.5. Scale (EMS)

In relation to the invariance of the measurement, this was assessed considering sex, taking into account the difference in the CFI (ΔCFI), and observing that this difference in the configural-metric, metric-scalar, and scalar-strict invariance is less than 0.010 and the difference in the RMSEA (ΔRMSEA) is less than 0.015, which shows that the eight items of motivation for sexting show adequate invariance in the measurement according to sex [66,67] (Chen, 2007; Dimitrov, 2010)

Table 4. Gender invariance of measurement in the motivation for sexting.

Modelo	χ2	Δχ2	gl	Δgl	P	CFI	ΔCFI	RMSEA	ΔRMSEA
Configural	19.05633		40		0.001	0.973		0.022
Metric	21.59985	−2.54352	47	−7	0.001	0.987	−0.014	0.018	0.004
Scalar	24.8401	−3.24025	54	−7	0.001	0.988	−0.001	0.017	0.001
Strict	29.16796	−4.32786	62	−8	0.001	0.989	−0.001	0.016	0.001

.

3.6. Reliability

Reliability was assessed through internal consistency using the entire one-dimensional model of the EMS, yielding a score of ω = 0.930.

4. Discussions

The general objective of this research was to validate a self-report instrument on motivations for sexting in Peruvian adolescents and young university students. The literature indicates that motivation is a key component in the formation and maintenance of habits and that it can directly influence behavior (Deci & Ryan, 1985; Bandura, 1986). In this context, the construct is defined as those attitudes or situations that motivate the population to send sexually explicit images through mobile devices to arouse sexual attraction or desire, which may be driven by specific characteristics or situations surrounding the individual [25,32,33,34,35,36,68,69] (Barroso et al., 2021; Chacón-López et al., 2019; Choi et al., 2019; Gassó et al., 2019; Klettke et al., 2019; Rodríguez-Castro et al., 2018; Seto et al., 2023; Van Ouytsel et al., 2020).

In the Exploratory Factor Analysis, the psychometric quality of the items was analyzed. Three items were eliminated under the criterion of double factorial loading since they showed strong factorial loadings on a dimension to which they did not correspond, demonstrating the poor discrimination of the items in reflecting the proposed dimension [64] (Reise et al., 2010). For the remaining 16 items, acceptable factorial loadings (>0.3) were obtained [65] (Ferrando & Anguiano-Carrasco, 2010).

For the confirmatory factor analysis, two fit models were constructed. In the first model, adequate results were obtained for the CFI and TLI, which resembled those obtained by [30] Chacón-López et al. (2016). However, the RMSEA = 0.106 score exceeded the appropriate parameters proposed by [55,70,71,72] MacCallum et al. (1996), Zhang et al. (2016), Lai (2021), and Xia & Yang (2019). Therefore, it was decided to remove eight items as they were not clear enough to represent the construct. This adjustment allowed for finding a unidimensional model of eight items that fit very well into Model 2, revealing better-fit indices for the CFI, TLI, and RMSEA, similarly to the results obtained in the construction and validation of the Sexting Behavior Scale (ECS), achieving scores that meet the parameters established by [55,70,71] MacCallum et al. (1996), Lai (2021), and Xia & Yang (2019). This showed that in the Peruvian context, different factors motivate sexting practice, which are expressed in the EMS, where it is observed that personal minimization and instability [14,37] (Bragard & Fisher, 2022; Madigan et al., 2018b), fear of losing one’s partner [73] (González et al., 2020), mismanagement of anxiety [74] (Dodaj et al., 2020), and peer or partner pressure [75,76] (Ochoa-Pineda, 2018; Paintsil et al., 2023) ultimately lead young people to engage in sexting.

Finally, for the reliability report, an adequate level was found according to the omega coefficient theory established by [77] Campo-Arias & Oviedo (2008), showing that the unidimensional model of the eight items demonstrates adequate internal consistency to measure the motivators of sexting, whose indicators are internal and external motivation [56] (Katz, 2007). These values resemble the Sexting Aptitude Scale [78] (Rodríguez-Castro et al., 2017). Although this method was not considered in the Sexting Motivators Questionnaire for Adolescents and Young People [79] (Bianchi et al., 2016), it is a widely used and recommended method for reporting the reliability of an instrument [80,81] (Watkins, 2018) (Watkins, 2018; Izquierdo-Cárdenas et al. 2021).

One of this study’s main limitations is the sampling method, as non-probabilistic sampling limits the possibility of generalizing the results. Similarly, some sociodemographic characteristics were imbalanced since the sample was drawn from a single region, considering individuals from both private and public universities. Therefore, it is recommended that future research consider probabilistic sampling to achieve a better distribution of participants. Another limitation was the cross-sectional data collection in only one region of the country. Although it is an extensive geographic region, the population size is not as large as the coastal region of Peru. Therefore, it is recommended that future research include other regions. The last limitation was that the scale was administered as a self-report, which could bias responses due to desirability. Therefore, it is suggested that future research use personal interviews instead.

Among the practical implications, this metric is available to the scientific community and health professionals, such as therapists, who could use it to evaluate the underlying reasons for sexting in their patients. In addition, by using the proposed metric, it will be possible to obtain diagnoses that allow for the design of educational programs that address sexting from informed perspectives, which would allow for understanding the motivations and risks associated with the practice of sexting.

In conclusion, the EMS demonstrates adequate psychometric properties for measuring motivations for sexting practice in university students. Furthermore, this research enables a more detailed analysis of the motivations behind this practice and facilitates the development of more specific intervention strategies to reduce them. In this context, future studies could apply the metric in longitudinal studies in order to discover the evolution of the motivations for sexting over time. Also, future research could integrate predictive models to identify any risk behaviors associated with non-consensual or coercive sexting.