📄 social_science

Academic Performance Trajectories Across Five Semesters: A Longitudinal Study of Undergraduate Students in a Philippine Higher Education Institution

This longitudinal study of 989 undergraduate students in a Philippine higher education institution reveals that while overall academic performance initially improves and then declines over five semesters, significant heterogeneity in these trajectories across different degree programs suggests that institution-wide averages may obscure critical program-specific trends, thereby supporting the need for longitudinal monitoring in academic advising and planning.

Original authors: Joelle Dianne Flores Venegas

Published 2026-09-01

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Joelle Dianne Flores Venegas

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the world of higher education, a student's success is often measured by a single number: their cumulative grade average. This figure acts as a summary, a final score that tells a school whether a student has passed or failed, qualified for a scholarship, or earned a spot on the honor roll. It is a convenient tool for administrators, but it has a blind spot. By averaging out every grade a student earns from their first day to their last, this single number can hide the actual story of their journey. Two students might finish with the exact same average, yet one could have started poorly and steadily improved, while the other began with high marks and slowly struggled as the years went on. These different paths matter deeply. Understanding whether a student is climbing or sliding, and when that shift happens, offers a much clearer picture of their experience than a static average ever could.

Researchers at a college in the Philippines set out to look past these summaries to see the actual movement of student performance over time. They gathered the academic records of nearly one thousand undergraduate students from ten different degree programs, tracking their grades across five consecutive semesters. Instead of looking at the final result, they watched how the numbers changed from one term to the next. In the grading system used by this institution, a lower number represents a better grade, meaning that if a student's score drops, they are actually doing better, and if the number rises, their performance is slipping. The study revealed that the overall group of students did not follow a straight line. Instead, their performance followed a distinct curve: students improved during their second semester, held steady through the third, and then began to decline in their fourth and fifth semesters.

This pattern of improvement followed by a later decline was not the same for every student. When the researchers broke the data down by major, they found that the story changed completely depending on which program a student was in. Some degree programs showed students getting stronger as they progressed, with grades improving significantly by the final semester. Others told a different tale, where students started strong but saw their performance drop sharply in the later years. For instance, students in one program saw their grades get better every step of the way, while students in another program experienced a noticeable drop in performance specifically in their final semester. The researchers found that the differences between these programs were so large that a single average for the entire college would completely miss these critical details.

The study suggests that looking at a student's recent history is just as important as looking at their total average. A student might have a decent overall score, but if their grades have been getting worse for the last two semesters, they might need help before they fail. Similarly, a specific degree program might be facing a hidden problem if students consistently struggle during a particular year, even if the program's overall reputation remains high. The researchers noted that they could not say exactly why these changes happened, as they only had access to grades and not to other factors like how much time students worked, their personal lives, or how difficult specific classes were. However, the clear message from the data is that academic life is a process, not just a final score. By watching how grades move over time, schools can spot trends earlier and offer support when it is actually needed, rather than waiting until a student has already fallen behind.

Technical Summary: Academic Performance Trajectories Across Five Semesters

Problem Statement
In higher education, academic performance is frequently reduced to a cumulative grade average (CGA) or a single point-in-time indicator. While convenient, this aggregation obscures meaningful longitudinal changes in student performance. Two students with identical CGAs may follow vastly different trajectories—one improving over time, the other deteriorating. Such differences are critical for academic advising, intervention, and curriculum evaluation but remain invisible in aggregate data. This issue is particularly salient in the Philippine context, where institutional records are organized by semester and degree program, yet aggregate averages often fail to reveal when performance shifts occur or whether different programs experience distinct trajectory patterns. Furthermore, the specific grading system used (where lower numerical values indicate better performance) requires careful interpretation of grade fluctuations as either improvement or deterioration.

Methodology
The study employed a quantitative, longitudinal, retrospective design using institutional academic records from a Philippine higher education institution.

Participants: The dataset included 998 undergraduate students enrolled in ten degree programs (e.g., BECED, BPA, BHS, BSIA, BSMA, BSSW, BSP, BSEMC, BSCrim, BPE). Of these, 989 students had complete academic records across five consecutive semesters, yielding 4,945 student-semester observations.
Variables: Academic performance was operationalized as the numerical semester grade. "Semester" was treated as a five-level within-student factor, and "Degree Program" as a categorical between-student factor.
Statistical Analysis:
1. Descriptive Statistics: Means and standard deviations were calculated for each semester and program.
2. Repeated-Measures ANOVA: Conducted on the 989 students with complete data to test for overall differences in performance across the five semesters.
3. Factorial Regression Model: A model including semester, degree program, and their interaction was estimated using student-clustered robust standard errors to account for repeated observations within students. Estimated Marginal Means (EMMs) were derived for each program-by-semester cell.
4. Post-Hoc Contrasts: Consecutive-semester contrasts were calculated with Holm adjustment for multiple comparisons to identify specific periods of significant change within and across programs.
- Significance was evaluated at $\alpha = .05$ .

Key Results

Overall Trajectory: Academic performance changed significantly across the five semesters ( $F(4, 3952) = 58.30, p < .001, \eta_p^2 = .056$ ). The trajectory was nonlinear:
- Semester 1 to 2: Significant improvement (Mean grade decreased from 1.503 to 1.437).
- Semester 2 to 3: Relative stability (Mean grade remained at 1.444).
- Semester 3 to 5: Significant deterioration (Mean grades increased to 1.491 in Semester 4 and 1.562 in Semester 5).
- The effect size was modest, indicating that while the change was statistically reliable, the average magnitude of change was relatively small.
Program-Specific Heterogeneity: A significant semester-by-program interaction was found in the regression model ( $F(49, 4895) = 83.45, p < .001$ ), explaining approximately 30.4% of the variance ( $R^2 = .304$ ). This confirms that no single institution-wide trajectory adequately represents all programs.
- Improvement: Programs like Human Services (BHS) and Psychology (BSP) demonstrated substantial improvement, particularly in later semesters.
- Deterioration: Public Administration (BPA), Information Administration (BSIA), and Social Work (BSSW) showed pronounced deterioration, especially in the final semesters. BSIA and BSSW exhibited particularly sharp deterioration in Semester 5.
- Fluctuation: Other programs, such as Early Childhood Education (BECED) and Criminology (BSCrim), displayed fluctuating patterns of improvement and deterioration across different semesters.

Key Contributions

Longitudinal Granularity: The study demonstrates that institution-wide averages conceal meaningful, program-specific trajectories. It provides empirical evidence that academic performance is a dynamic process rather than a static outcome.
Methodological Application: It applies repeated-measures ANOVA and factorial regression with clustered robust standard errors to educational records, offering a rigorous framework for analyzing student trajectories while accounting for within-student correlation.
Contextual Insight: It highlights the specific non-linear pattern (improvement $\rightarrow$ stability $\rightarrow$ deterioration) observed in this Philippine institution, challenging the assumption of linear progression or uniform decline.

Significance and Claims
The paper claims that longitudinal academic monitoring is essential for complementing conventional cumulative performance indicators.

For Advising: Trajectory information can identify students who are deteriorating even if their cumulative average remains acceptable, allowing for earlier intervention.
For Program Review: Consistent deterioration in specific semesters across a program (e.g., the Semester 5 deterioration in BSIA and BSSW) can serve as an initial signal for curriculum review or workload assessment.
For Institutional Planning: Retaining and analyzing student-level longitudinal records is more valuable than relying exclusively on aggregate summaries.

Limitations and Cautions
The author explicitly states that the study is observational and cannot establish causal mechanisms. The dataset lacks variables regarding attendance, socioeconomic status, motivation, or course difficulty. Therefore, the observed patterns cannot be attributed to specific causes (e.g., curriculum sequencing vs. student cohort characteristics) without further course-level analysis. The findings are specific to one institution and should not be generalized automatically to other contexts. The study does not propose specific interventions but rather advocates for the use of trajectory data in existing monitoring strategies.

Technical Summary: Academic Performance Trajectories Across Five Semesters

More like this