The research employed a quantitative, observational, hypothetical deductive, explanatory-level, non-experimental design, which is suitable for analyzing the relationships between variables without directly manipulating them (Creswell, 2014). The design was cross-sectional, as it collected data only once, with the aim of evaluating the impact of self-confidence on the development of digital skills, which are crucial in the current trend of digitalization in university education. The quantitative approach allows obtaining generalizable results from the analysis of numerical data and providing robust conclusions about the relationships between self-confidence and digital skills in university students. As noted by Hernández et al. (2014). In this case, the aim is to predict how self-confidence impacts the development of digital skills.

The sample of 213 university students was selected through stratified sampling, ensuring adequate representation of diverse geographic areas and socioeconomic levels. “Stratified sampling provides a proportional representation of different subgroups in a population” (Cohen et al., 2018). This approach enabled the capture of contextual differences in terms of access to technology and self-confidence, which are important factors in the development of digital skills. Furthermore, considering that Peruvian university students face significant barriers to accessing advanced technologies, particularly in rural areas (Mendoza, 2023), this type of sampling is crucial to capture these variations. The chosen sample, by region, included 33.8% from the mountains, 31.9% from the jungle, and 34.3% from the coast; the ages ranged from 23 to 63 years, with 27.7% between 38 and 42 years old (see Table 1).

The main variables of the study are:

Self-confidence: It is defined as the “belief in one's own ability to carry out a task or face a challenge successfully” (Bandura, 1997, 1986). To measure this variable, the Academic Self-Confidence Scale (Jones, 2020) was used, which has been shown to be valid and reliable in previous studies on digital skills in higher education.

Digital skills: These are defined as “the ability to use digital technologies in a critical, creative, and efficient way” (European Commission, 2018). The Digital Skills Questionnaire for University Students (Fernández and Ruiz, 2020) enabled the evaluation of multiple dimensions, including the technical management of digital tools and the capacity for critical analysis.

Educational digitalization: The study evaluated students' ability to use digital tools for academic task development, a key dimension. Given the growing importance of digitalization in our country and its demonstrated versatility and usefulness in virtual education, it is crucial for educational innovation in the 21st century (Pérez et al., 2022; Rosales Márquez, 2022).

The following instruments were used:

The data collection process was carried out in the second semester of 2023. As part of the collection procedure, an online platform was used that facilitated remote access to the questionnaires by students, especially in regions with lower connectivity. Statistical authors (Cohen et al., 2018) indicated the usefulness of digital surveys to facilitate collection in samples with dispersed geography.

In addition, in the semi-structured interviews with teachers, a question guide was used focused on identifying difficulties in using technological tools as part of the digitalization of education, particularly in the lack of self-confidence in students. The interviews were recorded and transcribed for a subsequent qualitative analysis, enabling a more comprehensive understanding of the structural barriers. As noted by Creswell (2014), “the qualitative mixed approach allows contextualizing the numerical findings” (p. 232).

Data analysis was performed using SPSS software (version 26.0). Initially, descriptive analyses were carried out to characterize the sample. Subsequently, Pearson's correlation test was used to assess the relationship between self-confidence and digital skills. This technique effectively measures the degree of association between two continuous variables, enabling the assessment of the relationship's strength and direction (Hernández et al., 2014).

On the other hand, for predictive analysis, regression was performed using Jamovi 2.3.28 software, which predicts the value of a dependent variable based on one or more independent variables (Field, 2013). Likewise, statistical mediation analysis was used, which is particularly useful in this study, as it allows estimating the impact of self-confidence on the development of digital skills while controlling other variables such as age, gender, and socioeconomic context (Cohen et al., 2018).

The Table 2 displays the correlation coefficient matrix between university students' self-confidence and their digital competence. Correlations with coefficients ranging from (0.313 to 0.822) are found, indicating moderate to strong, positive, and highly significant relationships (p < 0.001). These results show that after verifying the correlation between the study variables' dimensions, the first phase of the explanatory analysis begins.

The Table 3 shows the bivariate relationship between the variables self-confidence and digital skills. It reveals a highly significant moderate relationship of a direct type at the dimension level, indicating that as self-confidence increases, so does the level of digital competence.

The analysis demonstrates the predictive capacity of self-confidence in the digital skills of Peruvian university students (Table 4). Evidence confirming a causal relationship between self-confidence (predictor) and digital skills (predicted) suggests that an increase of one unit in self-confidence leads to an expected increase of 0.377 units in the digital skills of university students, implying a positive correlation between self-confidence and improved digital skills. The confidence interval reinforces the fulfillment of a positive relationship between the variables, with a significance < 0.001.

The regression Table 5 shows the predictive capacity of the dimensions of self-confidence with respect to digital skills. Each row represents one of these variables, and the numbers associated with it indicate its importance and relationship with digital competence. Significant values were evidenced for the predictors: study (0.385); verbalization (0.011); marginally significant predictors like compression (0.371) and student grades (0.247), as well as non-significant predictors such as clarification (0.803). All these dimensions, simultaneously serving as predictor variables, demonstrated varying degrees of significance in elucidating the development of digital skills among university students, with the study and verbalization dimensions garnering the most importance.

The diagram enables the examination of the mediating role of sex (male or female) in the direct relationship between self-confidence and digital skills in university students. It reveals that the mediating or indirect effect (−0.00) (−0.24) holds less significance than the direct relationship between both variables (0.38); consequently, sex does not serve as a mediating effect, as it does not enhance the relationship between the variables under investigation (Figure 1).

The indirect effect on digital skills is 1.58E-04; this effect reveals the mediating role of gender. However, the confidence interval includes zero, suggesting that this effect is not statistically significant. On the other hand, the direct effect of self-confidence on digital skills is 0.377 and is statistically significant (p < 0.001). This indicates that, independently of any mediating effect, self-confidence has a direct influence on digital skills. The total effect of self-confidence on digital skills is 0.377, which is equal to the direct effect. This confirms that the direct effect primarily explains the relationship between gender and digital skills, while the indirect effect is not significant (Table 6).

Figure 2 illustrates the impact of age as a mediating function on the direct relationship between self-confidence and digital skills in university students. It reveals that the mediating or indirect effect (−0.03) (−0.06) has a lower statistical significance compared to the direct relationship between both variables. Consequently, age did not demonstrate a mediating effect that enhances the relationship between self-confidence and digital skills.

The table shows that there is an indirect effect on digital skills of 0.00193 with a confidence interval that includes zero, which proves that this effect is not statistically significant. The direct effect on digital skills is 0.37526 and is statistically significant (p < 0.001). This indicates that, independently of any mediating effect, self-confidence has a direct influence on digital skills. The total effect of self-confidence on digital skills is 0.37720, which is close to the direct effect. This demonstrates that the direct effect between the variables under study primarily explains the relationship between sex and digital skills and that the indirect effect is not significant.

The region of origin of the students does not appear to have a mediating effect on the direct relationship between self-confidence and digital skills, which are crucial to develop given the impending digitalization of education (Table 7).

The table indicates that the direct effect holds a higher statistical significance (0.48325) (p < 0.001). In contrast, the analysis failed to demonstrate the indirect effect that determines the mediation. This means that the condition of coming from a specific region of Peru does not alter the relationship between both variables (Table 8).

This analysis collected the opinions of university teachers regarding the relevance of developing digital skills in students so that they can guarantee their optimal academic performance in their higher education experience, from which the following comments on the subject were compiled:

The predictive capacity of self-confidence in digital skills is key to projecting future trends in the digitalization of education in Peru. As University Law 30220 encourages the total virtualization of university classes, it is expected that students with greater self-confidence will be the first to adapt to the new educational dynamics. Self-efficacy theories (Bandura, 1997) suggest that individuals who develop strong self-confidence in their abilities tend to maintain and improve these skills in the long term (Pajares and Schunk, 2001). In this sense, it could be predicted that Peruvian students with high levels of digital self-confidence will be the first to integrate emerging technologies such as artificial intelligence into their learning.

Self-efficacy theory (Bandura, 1997) is key to explaining the results obtained and projecting the future impact of self-confidence in the educational field. People who believe in their abilities to complete complex tasks, such as using digital tools, will be more successful in facing the challenges of digitalization (Luszczynska et al., 2005). This theory can help predict that, in a scenario with greater access to digital tools and better technological inclusion policies, self-confidence will continue to be a crucial factor in academic success (Serdyukov, 2017; Shunk and Pajares, 2010) also supports this idea by suggesting that digital literacy, mediated by self-confidence, is essential for continuous learning in the digital age.

The correlation between self-confidence dimensions and digital skills is moderate to strong, with coefficients between 0.313 and 0.822 (Table 2). These results are in line with global research showing that self-confidence in learning by themselves plays a fundamental role in the development of technological and other skills (Rosales-Márquez et al., 2023). At the national level, the Ministerio de Educación (2023) reports that students with greater self-confidence in their technological skills tend to adapt better to advanced digital tools. The achievement motivation theory, which posits that two factors drive individuals to achieve goals: the desire for success and the fear of failure, finds confirmation in these results. Students who have high levels of self-confidence tend to have a greater motivation to achieve digital skills, as they perceive the use of technologies as an achievable challenge (Atkinson, 1964). This motivation may explain the positive relationship between self-confidence and the development of digital skills.

Including activities that encourage reflective practice in educational programs could help students internalize their achievements and learn from their mistakes, thereby increasing their self-confidence. For example, by reviewing their use of digital technologies, including AI, students could develop a deeper understanding of their capabilities and limitations, progressively improving their digital skills. International studies, such as those by Field (2013), have also shown that reflection on digital practice promotes more efficient and effective use of technologies in educational contexts.

Mediation analysis of sex in the relationship between self-confidence and digital skills suggests that sex does not have a significant mediating effect (Figure 1). This is in line with international studies, such as that by Pérez et al. (2022), who found that sociodemographic factors such as gender have a minor impact on the adoption of technologies, while self-confidence remains the main predictor of digital skills.

The digital literacy theory (Eshet-Alkalai, 2004) emphasizes the importance of advanced cognitive skills for handling technological tools, and Table 4 shows that predictors like “verbalization” and “studying” have a greater influence on digital skills. Another author emphasizes the importance of verbalization to generate greater self-confidence and greater acquisition of digital skills (Chávez and Montes, 2015). In turn, Figure 2, which examines the mediating effect of age, can be analyzed under the digital divide theory (van Dijk, 2006), since older students may have had less access to technology in their previous training, which influences their current confidence in using digital tools. Figure 3 examines the unproven mediating effect of the region of origin. This can be justified by the educational level since, in Peru, the universities are located in central cities in the different regions and departments of the country, where students can access technology, so their region of origin did not make a difference between the groups.

The results of the study confirm that self-confidence is a significant predictor of the development of digital skills in Peruvian university students. A moderate positive correlation (r = 0.484, p < 0.001) was found between both variables, indicating that students with higher levels of self-confidence are more likely to develop advanced digital skills. This finding highlights the importance of fostering self-confidence in educational settings, as students who are confident in their ability to use digital tools tend to use emerging technologies, such as artificial intelligence (AI), more effectively, which improves their academic performance and their ability to innovate in the context of the digitalization of university education.

The study also reveals that a lack of self-confidence perpetuates structural barriers that limit the access and adequate use of digital technologies in university contexts, especially in regions with lower connectivity and resources. The results show that self-confidence directly influences the effective use of digital technologies, as students with low levels of self-confidence tend to avoid or underuse these tools (β = 0.377, p < 0.001), which exacerbates existing digital gaps. These results suggest that, in order to close technological gaps in Peruvian higher education, it is crucial to implement pedagogical strategies that not only provide access to technology but also develop self-confidence in the use of these tools.

The analysis of the quantitative data in contrast to the quantitative data shows us that for the digitization of higher education, first of all, self-confidence must be developed to take on new challenges, as suggested by the one of digital platforms, AI, software, among others. In this sense, it is essential that students and teachers have developed digital skills oriented to education. The university should foster self-confidence in its students through interventions such as workshops, and teachers should play a strategic role by accompanying them and providing trust, enabling them to tackle new challenges, including those posed by digitized education.

To strengthen the development of digital skills in Peruvian university students, it is recommended to implement training programs focused on increasing self-confidence in the use of emerging technologies, such as artificial intelligence. These programs should include practical workshops and mentoring that promote the critical and creative use of digital tools, allowing students to become familiar with their operation and apply them in academic tasks. In addition, it is crucial that universities adapt their digital inclusion policies, ensuring that both students and teachers have the necessary support to overcome technological barriers, especially in regions with fewer resources.