Introduction
In recent decades, academic migration has emerged as a pivotal trend in higher education, with students increasingly pursuing studies abroad. In Iran, approximately 23,000 international students are enrolled, predominantly from Afghanistan (Mousavi, 2019). UNESCO data from 2013 indicates that around 17,000 Afghan students were studying overseas, with 53% enrolled in Iranian institutions (Staff, 2016). Given the unique challenges migrant students face in foreign academic environments, understanding their self-concept is critical. Self-concept, encompassing an individual's positive or negative beliefs about themselves, significantly influences their social and psychological adjustment (Hosseinian, Ghasemzadeh, Anousheh, & Rasouli, 2014). Factors such as family dynamics, educational settings, peer feedback, and cultural values shape self-concept (Abouei, 2019).
Quality of life, a multifaceted construct, includes physical, psychological, social, spiritual, and health-related dimensions, influenced by social interactions and personal competencies (Rajabli, 2019). Research highlights a connection between self-concept and quality of life (Vaquero-Solís et al., 2021; Mirakbarzadeh, Khanjarkhani, & Nastizaei, 2020; Sanei & Raeisoun, 2020). Similarly, social skills—behaviors that foster positive interpersonal relationships—impact both self-concept and quality of life (Daneshfar, 2013; Seraj Amani, 2015).
While prior studies have explored the relationships between self-concept and quality of life or self-concept and social skills independently, no research has concurrently examined the interplay of these three variables among Afghan students. This study addresses the question: How does self-concept influence the quality of life and social skills of Afghan students at Birjand University?
This study tests the following hypotheses:

Self-concept is significantly associated with quality of life.
Self-concept is significantly associated with social skills.

Self-concept is significantly associated with both quality of life and social skills.
 Research Method
This study utilized a correlational research design with a convenience sampling approach. The population comprised all Afghan students enrolled at various academic levels at Birjand University, totaling 470 students. Based on predefined inclusion and exclusion criteria, 212 Afghan students participated in the study. The Tennessee Self-Concept Scale is a 100-item questionnaire employing a 5-point Likert scale (Jamaluddin, Ahmad, Yusuf, & Abdullah, 2009, cited in Safaei, 2010). It has demonstrated robust validity and reliability across studies, with reported reliability coefficients of 0.92 (Hines, 1989), 0.82 (Rezazadeh Bahadoran, cited in Chenari, 1995), 0.74 (Mahmoudi, 1998), 0.80 (Houshmand, 1997, cited in Safaei, 2010), 0.83 (Moradi, 2002), and 0.98 (Rabiei-Nejad, Kajbaf, Mazaheri, Talebi, & Abedi, 2015). The Quality-of-Life Questionnaire consists of 26 items across four domains. Reliability coefficients for each domain are as follows: Physical Health (0.77 and 0.75), Psychological Health (0.77 and 0.76), Social Relationships (0.75 and 0.70), and Environmental Health (0.84 and 0.79) (Nejat, Montazeri, Holakouie, et al., 2006, cited in Khalilzadeh Behrouzian & Ahmadi, 2013). The Inderbirtzen and Foster Social Skills Questionnaire had 39-item questionnaire features six response options and includes two subscales: positive behaviors and negative behaviors. The overall reliability is 0.84, with subscale reliabilities of 0.90 for positive behaviors and 0.72 for negative behaviors. The internal consistency is 0.88, and the questionnaire exhibits satisfactory convergent and discriminant validity (Daneshfar, 2013).
 Findings
To test the research hypotheses, multivariate regression analysis was conducted using the simultaneous method. The assumptions of univariate and multivariate normality, linearity, homogeneity of variance, and absence of multicollinearity were verified and met, allowing the regression analysis to proceed. The results for the three hypotheses are as follows:

Hypothesis 1: Quality of life components explained 29% of self-concept variance.
Hypothesis 2: Social skills components explained 31% of self-concept variance.
Hypothesis 3: Quality of life and social skills components together explained 45% of self-concept variance.

All three regression models were statistically significant (F-tests, p<0.001), confirming a significant relationship between the predictor variables and self-concept. Analysis of regression coefficients showed:

For Hypothesis 1, all quality-of-life components were significant (t-tests, p<0.05).
For Hypothesis 2, all social skills components were significant (t-tests, p<0.05).
For Hypothesis 3, physical health, social relationships, and negative behaviors were significant (t-tests, p<0.05).

Only statistically significant components meaningfully contributed to predicting self-concept in each model.
 Conclusion
The findings confirm all the three hypotheses of the study. Self-concept is significantly related to quality of life. Self-concept is significantly related to social skills. And finally, self-concept is significantly related to both quality of life and social skills combined.