Analysis of academic performance from a binary logistic regression model

M. Pérez, O. Mejía, C. Serrano, S. Suescún-Garcés, O. Mogollón-Alaguna, F. León

Research output: Articles / NotesScientific Articlepeer-review

Abstract

Introduction: Comprehending the factors influencing students' academic performance holds significant importance for the Universidad de Santander. This understanding enables the university to implement curricular adjustments and adaptations, which play a fundamental role in fostering the development of student competencies. Consequently, these adjustments contribute to enriching educational processes, thereby aiding in the successful fulfillment of the educational system's mission. Objective: To examine the factors influencing the academic performance of incoming students as opportunities for improvement that address the students' needs. Materials and Methods: This study adopts an exploratory and cross-sectional design. Its comprised 1,161 new students. The response variable under consideration is the academic average attained by students at the conclusion of the academic semester. Data were sourced from national educational tests and institutional information systems. It was performed a statistical analysis using binary logistic regression, employing SPSS version 26 for the statistical software. Results and Discussion: An analysis of variance ANOVA F (2) = 24.94, p<.001 was performed, finding significant differences between the means of the average in the three campuses. The bivariate analysis using the Χ2(2) test = 26.72, p<.001, indicates that there is a statistically significant association between the academic average and the campus to which the students belong. Furthermore, the competencies assessed by the Saber 11 test, particularly the performance levels achieved in English and mathematics, were identified as crucial factors for the estimation of the academic performance model through binary logistic regression. Conclusions: Students who enter college with a stronger foundation in mathematics, critical reading, citizenship, and English proficiency experience enhanced consolidation within the college teaching and learning environment.

Translated title of the contributionAnálisis del rendimiento académico a partir de un modelo de regresión logística binaria
Original languageEnglish
Pages (from-to)1-14
Number of pages14
JournalInnovaciencia
Volume11
Issue number1
DOIs
StatePublished - 2023
Externally publishedYes

Keywords

  • Early warning systems
  • Professional competence
  • Quality improvement
  • Regression analysis
  • School underachievement

Fingerprint

Dive into the research topics of 'Analysis of academic performance from a binary logistic regression model'. Together they form a unique fingerprint.

Cite this