In recent decades, academic goals have been consolidated as fundamental orientators of students' efforts to improve their performance (Mendo-Lázaro et al., 2022). This motivational construct is articulated with the metacognitive skills essential for planning, monitoring and evaluating one's own learning process (Rahman et al., 2024). However, there is still a lack of sufficiently reliable instruments to measure these competencies in current educational environments (Nájera et al., 2020). This lack hinders alignment with SDG 4, which promotes inclusive, equitable and quality education, as curricular rigidity, institutional resistance and lack of funding and teacher training limit its effective implementation (Ferguson and Roofe, 2020; Zickafoose et al., 2024).

Although the Ministry of Education has incorporated participatory methodologies to define and monitor academic goals in higher education, there is still an urgent need to design and validate a specific instrument for university students in Lima (Jara et al., 2022). Closing this gap is essential to strengthen academic engagement, coping with academic stressors, improving academic skills, and achieving academic achievement and success in various academic contexts (Aladini et al., 2024). Likewise, having a validated instrument will facilitate the improvement of the university experience and the incorporation of values such as competitiveness and performance, favoring an excellent education in marginalized contexts (Oliveira et al., 2025; Sloan-Lynch and Morse, 2024).

Bandura (1986) conceptualized academic goals as concrete intentions that channel behavior toward formative achievements, integrating them into his Social Cognitive Theory as key tools of self-regulation. Subsequently, Lent et al. (1994) adapted this framework to the university setting through the Social Cognitive Career Theory (SCCT), highlighting three interrelated components: academic goals, self-efficacy beliefs and outcome expectations, whose interaction determines student planning, perseverance and performance.

Based on the SCCT of Lent et al. (1994), it is postulated that college motivational dynamics are influenced by self-efficacy beliefs, outcome expectations and personal goals. From this foundation, Lent et al. (2007) developed several scales, among them, the Academic Progress Goals Scale, a seven-item unidimensional scale that assesses students' perceived progress in achieving their academic goals, i.e., the degree to which they feel they are making progress toward goals such as “learning and understanding the material in each of their courses” (p. 432). This instrument reported good internal consistency (Cronbach's α = 0.86).

To ensure an accurate assessment of the construct, it is essential to operationally define each type of goal, since the motivational process depends on the recognition of aspirations, obstacles and self-perception of one's own performance (Sheu et al., 2010). Therefore, it is proposed to adapt the original scale to correct conceptual ambiguities and improve its applicability in different contexts. This adaptation is based on the theoretical model of Lent et al. (2007), whose robustness allows for a valid assessment of academic goals and the development of interventions aimed at academic and vocational strengthening.

Within the conceptual framework, academic goals symbolize the future objectives that students seek to achieve, conditioning their involvement in specific activities (Bandura, 1986; Zalazar-Jaime and Cupani, 2018b). Two main types are distinguished: (a) Choice goals: linked to the domain or area that the student aspires to master; (b) Performance goals: related to the level of performance that the student wishes to achieve.

Both goals play differentiated roles: one motivates the exploration of educational options; the other drives the achievement of performance standards in specific tasks (Zalazar-Jaime and Cupani, 2018a). Taken together, they support behavioral self-regulation and are associated not only with satisfaction in the academic context, but also with general subjective wellbeing (Lent et al., 2007). Numerous studies have demonstrated their relevance in interaction with variables such as self-efficacy and outcome expectancies (Işik et al., 2018; Sheu et al., 2022; Sheu and Bordon, 2016).

The Lent et al. (2007) scale has been widely used in several countries such as: Spain (Lent et al., 2017), Portugal (Lent et al., 2009), Angola and Mozambique (Lent et al., 2014), Italy (Lent et al., 2011), Argentina (Medrano et al., 2017), Turkey (Işik et al., 2018) and USA (Duffy and Lent, 2009; Hui et al., 2013; Lent et al., 2007) with acceptable internal consistency (α and ω >0.80). However, so far there is only one research that explored its internal structure (Zalazar-Jaime and Cupani, 2018a), finding satisfactory results (CFI.993, TLI.990, RMSEA = 0.072). Therefore, the need emerges to examine in depth the psychometric properties of the Academic Progress Goals Scale, providing solid empirical evidence for its application in various contexts.

The availability of an accurate instrument to measure academic goals in higher education responds to the interest of promoting learning and optimizing the development of student capabilities (Valle et al., 2006). Having an adapted tool contributes to: (a) strengthening research actions and curriculum improvement (Compagnucci and Spigarelli, 2020; Fernández-Bringas et al., 2024; Pérez et al., 2017); (b) identifying areas for improvement in competencies such as critical thinking, creativity, cognitive flexibility and comprehension(Gaviria and Corredor, 2024; Gökçe and Güner, 2024; Rebecchi et al., 2024; Rodríguez Pulido et al., 2021; Weng et al., 2025).

In addition, some studies have linked goal achievement with the behavioral regulation and organization necessary for academic success (Valle et al., 2009), as well as with the explanation of academic satisfaction through outcome expectations (Lent, 2004; Lent et al., 2007). Moreover, motivation, influenced by goals, impacts cognitive, affective and behavioral domains (Closas et al., 2011; Gaeta et al., 2015). In work contexts, the pursuit of career goals does not always lead to satisfaction, which highlights the differences between academic and professional environments (Abraham et al., 2024; Brown and Lent, 2020).

This study is justified by the urgent need to validate an instrument to measure academic goals in university students. The relevance of this measurement is evidenced by the high dropout rates: in Latin America, almost half of the population aged 25–29 did not complete their studies (Seminara, 2021), and in Peru the dropout rate was 12.6% in 2023 (Ministerio de Educación del Perú, 2021). Although the socioeconomic factor stands out as the main cause (Superintendencia Nacional de Educación Superior Universitaria, 2021), there are also elements related to performance, individual characteristics, family environment, academic adaptation and vocational orientation (Bardales et al., 2025; Quincho et al., 2024; Seminara, 2021).

In this framework, academic goals, understood here as a set of intrinsic and extrinsic factors linked to the educational environment, could facilitate the understanding of the causes of dropout, whose consequences include social and economic problems and underemployment (Aisa et al., 2019; Jia and Ericson, 2017; Silva-Laya et al., 2020), increasing inequality gaps. Having a validated and reliable instrument would contribute to mitigating these rates, foster subjective wellbeing, raise academic satisfaction and reinforce learning objectives.

Consequently, the main objective of this work is to validate an instrument to measure academic goals in higher education students, providing empirical evidence on its factorial structure and reliability, and offering a useful resource for researchers, teachers and educational policy makers.

Consistent with the findings presented, the main objective of this paper is to validate an instrument to measure academic goals in higher education students, providing empirical evidence on its factorial structure and reliability, and offering a useful resource for researchers, teachers and educational policy makers. Based on the theoretical foundation previously exposed, in recent decades academic goals have been consolidated as fundamental orientators of students' efforts to improve their performance (Mendo-Lázaro et al., 2022). Thus, this motivational construct is articulated with the essential metacognitive skills to plan, monitor and evaluate one's own learning process (Rahman et al., 2024). However, there is still a lack of sufficiently reliable instruments to measure these competencies in current educational settings, which hinders alignment with SDG 4 and limits the implementation of inclusive, equitable and quality education (Ferguson and Roofe, 2020; Nájera et al., 2020; Zickafoose et al., 2024). Similarly, this lack prevents the systematic comparison of pedagogical interventions at the international level, making it difficult to identify effective practices that could be replicated in the Peruvian context.

Consequently, in Latin America and, in particular, in the Peruvian context, the need for solid and contextualized psychometric tools persists (Aliaga, 2021; Nájera et al., 2020). In fact, the absence of a validated instrument makes it difficult to detect and mitigate inequalities in the achievement of academic goals, which has repercussions on high university dropout rates (12.6% in 2023) and regional inequity (Aladini et al., 2024; Ministerio de Educación del Perú, 2021; Oliveira et al., 2025; Sloan-Lynch and Morse, 2024). Therefore, the Ministry of Education has promoted participatory methodologies (Jara et al., 2022), but a specific instrument for university students in Lima that responds to local cultural and linguistic characteristics is still required. In addition, this instrument will allow evaluating the effectiveness of tutoring and mentoring programs, as well as student retention policies, providing baseline data for the continuous improvement of such strategies.

From a practical standpoint, the five-item AGPS can be administered in less than one minute and therefore lends itself to multiple points of contact within the university support ecosystem. It can be embedded in initial advising forms or orientation interviews to flag students who perceive limited progress, incorporated into learning-management systems and mid-semester surveys as part of early-alert dashboards, or used during one-to-one counseling to inform personalized action plans and referrals to tutoring or wellbeing services. Because of its brevity, the scale can also be re-administered later in the term, allowing advisors to track change over time and adjust interventions in real time.

Based on the Social Cognitive Career Theory (SCCT) of Lent et al. (1994), which integrates academic goals, self-efficacy beliefs and outcome expectations, it is postulated that college motivational dynamics depend on the interaction of these components. In this sense, Bandura (1986) conceptualized academic goals as concrete intentions that channel behavior toward formative achievement, and Lent et al. (2007) developed the Academic Progress Goals Scale, a seven-item unidimensional instrument with good internal consistency (α = 0.86). However, its applicability in Peruvian contexts had not been explored, which justifies the present adaptation. To ensure semantic, conceptual and experiential equivalence, each item was reviewed by two bilingual experts and tested with students from the coastal and Andean regions. This process acknowledged the regional variants of Spanish and the collectivist framing of academic success, often experienced as a familial responsibility among first-generation students, thus ensuring that the scale truly resonates with local motivational meanings.

To address this shortcoming, our study implemented an empirical and theoretically grounded adaptation of the original scale, correcting conceptual ambiguities and improving its applicability in diverse academic contexts. Likewise, the introduction of Exploratory and Confirmatory Factor Analysis procedures optimized by means of artificial intelligence represented a relevant methodological innovation (Goretzko and Bühner, 2022; Rodriguez-Rodriguez et al., 2025). In this way, we sought not only to confirm the unidimensional structure, but also to ensure the cultural and linguistic relevance of the instrument. This hybrid approach of traditional psychometrics and artificial intelligence opens new possibilities for the validation of scales in environments with limited technical and human resources.

In descriptive terms, students showed a predominant tendency toward “Good progress” in items MPA1 (62%) and MPA7 (59%), while item MPA4 stood out in “Excellent progress” (32%). In contrast, intermediate categories such as “Fair progress” were concentrated in MPA3 (29%) and MPA6 (22%), which could reflect individual differences in metacognitive skills or specific academic demands. These findings underscore the need for differentiated interventions that enhance learning regulation skills, particularly in those students who perceive moderate progress.

Regarding the internal structure, the AFE indicated that a one-factor configuration with five items (AGPS1, AGPS3, AGPS4, AGPS6 and AGPS7) offers an optimal fit. Specifically, iterative AI eliminated items with insufficient loadings and evaluated comparative and absolute indices, optimizing the model. Next, the CFA confirmed this structure with excellent fit indices, which is consistent with previous evidence in other contexts (Zalazar-Jaime and Cupani, 2018). Thus, it is demonstrated that the scale can be simplified from seven to five items without losing validity, facilitating its administration in environments of high academic demand.

In relation to reliability, the omega coefficient was used, which is more appropriate for congeneric models (Savalei and Reise, 2019; Ventura-León and Caycho-Rodríguez, 2017). The results yielded an average value of ω = 0.85, exceeding the threshold of 0.70 and aligning with international studies reporting α and ω > 0.80 (Lent et al., 2007; Medrano et al., 2017). Therefore, the scale consistently measures different aspects of academic progress goals. This level of reliability allows its use in both longitudinal research and one-off assessments, providing methodological flexibility to users.

Additionally, using a bootstrap procedure of 1,000 replicates in R, the stability of the factor structure was evaluated. This analysis revealed that the CFI and TLI indices maintained mean values of 1.00 (SD = 0.00), while the SRMR and CRMR remained below 0.03. Although the RMSEA showed greater variability (mean = 0.04; SD = 0.03), only 6 % of the replicates exceeded the cutoff point of 0.08. Thus, conclusive evidence was provided on the replicability and structural validity of the model (Awang et al., 2015). These results support the robustness of the instrument, showing that it is resistant to sampling fluctuations and suitable for different cohorts of students.

On the other hand, concurrent validity, evaluated by means of structural equation modeling, confirmed a strong association between academic progress goals and academic satisfaction. This finding supports the proposal of (Lent et al., 2007) and agrees with previous studies in students and teachers (Hui et al., 2013; Lent et al., 2011), evidencing that goal attainment promotes the satisfaction of psychological needs and general subjective wellbeing (Diener, 1984; Klug and Maier, 2015).

However, this study has some limitations. First, non-probability sampling restricts the generalizability of the results and may introduce sampling bias, given the over-representation of female students (64%) and health-science majors (55%). Future work should replicate the study using stratified probability designs or apply post-stratification weights to verify the stability of the findings. Furthermore, although the validated scale retains five items for the factor, future work could explore the incorporation of additional items to enrich the measurement (Brown, 2015). Second, the AI component was limited to producing brief conceptual rationales after a purely statistical, iterative algorithm had determined item retention or deletion; therefore, ChatGPT played no role in the pruning decisions themselves. While two psychometric reviewers screened every rationale, the possibility of incomplete or hallucinated explanations remains, and these narratives should not be regarded as definitive theoretical evidence. Third, because the optimal-deletion algorithm was calibrated on a convenience sample of Peruvian university students, it may capitalize on sample-specific variance; replications with probability samples and in other cultural-linguistic contexts are needed before generalizing the factorial solution. Finally, it would be valuable to extend the sampling to different regions and the social strata of the country to strengthen external validity. It is also suggested to investigate the sensitivity of the scale to specific teacher training interventions and tutoring programs, in order to evaluate changes at the level of academic goals over time.

In summary, the validation of the Academic Progress Goals Scale provides a valid, reliable and culturally adapted tool to the Peruvian university context. The results confirm a solid unidimensional structure, high internal consistency (ω = 0.85) and significant relationships with academic satisfaction, supporting its usefulness for identifying motivational profiles and designing pedagogical interventions. Despite sampling and virtual collection limitations, this study lays the groundwork for future research with more representative designs and strengthens the application of SCCT theory in improving student performance and wellbeing. Ultimately, having this adapted instrument contributes to the fulfillment of SDG 4 and the advancement of quality, equity and inclusion in Peruvian higher education, thus fulfilling a central purpose for this work.