The COVID-19 pandemic accelerated the shift to online learning in higher education. While some research suggests online learners underperform compared to in-person learners, a comprehensive examination of its effects on student achievement remains limited. This study aims to fill this gap by comparing the final exam results of 8297 freshmen in 2020 (in-person) and 8425 freshmen in 2021 (online) across five core courses at King Saud University, Saudi Arabia. These courses – University Skills 101, Entrepreneurship 101, Computer Skills 101, Computer Skills 102, and Fitness and Health Culture 101 – were chosen to represent a variety of subject matter and course structures. The study accounts for confounding variables like gender, class size, and admission scores, known to influence student performance. The rapid adoption of online learning during the pandemic, coupled with the growing use of technology in education, necessitates a thorough understanding of its effects on student outcomes to inform effective pedagogical strategies and policy decisions. The university's transition to online learning provides a unique case study to examine this impact.

Existing research on online versus in-person learning shows mixed results. Some studies consistently report lower achievement for online students (Fischer et al., 2020; Bettinger et al., 2017; Edvardsson and Oskarsson, 2008), while others find online students outperforming their in-person peers (Iglesias-Pradas et al., 2021; Gonzalez et al., 2020). Many studies emphasize the importance of student characteristics (Chesser et al., 2020; Cacault et al., 2021) and instructor effectiveness (Bergstrand and Savage, 2013) as influencing factors in academic achievement regardless of the learning modality. The theoretical framework for this study draws upon constructivism, which emphasizes the learner's active role in knowledge construction, and subject-specific pedagogy, which highlights the need for teaching methods tailored to the specific characteristics of each discipline. These theories suggest that effective instruction, whether online or in-person, should encourage active engagement and align with the subject matter's inherent nature.

This quantitative study analyzed final exam scores from 16,722 freshmen at King Saud University. The data included scores from 8297 students in in-person classes in 2020 and 8425 students in online classes in 2021. The five courses involved varied in credit hours (one to three) and learning activities (lectures and labs). The study controlled for potential confounding variables—gender, class size, and admission scores—using multiple linear regression analyses. Independent-samples t-tests were initially used to compare mean scores between the two years. The admission scores were obtained from the Deanship of Admission and Registration Affairs at King Saud University, representing a composite measure of students' high school grades and college entrance exam performance. Data on class size and gender distribution was also obtained from the same source. Reliability of the final exams was assessed using Cronbach's alpha (ranging from 0.77 to 0.86), demonstrating good internal consistency. Content, criterion-related, and construct validity were also established through expert reviews and correlation with admission scores. Ethical approval was obtained from King Saud University's Research Ethics Committee.

Initial t-tests showed statistically significant lower scores in online courses (2021) compared to in-person courses (2020) in all courses except Computer Skills 101. However, after controlling for gender, class size, and admission scores via multiple regression, a different pattern emerged. In University Skills 101, Fitness and Health 101, and Computer Skills 101, the 2021 online cohorts outperformed their 2020 in-person counterparts by 0.89, 0.56, and 5.28 points respectively. Conversely, the in-person group outperformed the online group in Entrepreneurship 101 by 0.69 points. No statistically significant difference was found in Computer Skills 102. The regression models explained a moderate amount of variance in final exam scores (adjusted R² ranging from 0.23 to 0.78 depending on the course).

The study's findings challenge the simplistic notion that online learning is inherently inferior to in-person instruction. The significant differences between the t-test and regression results underscore the importance of controlling for confounding variables. The modest differences observed in some courses (less than one-point difference) suggest that for introductory-level courses, the mode of instruction may have limited impact provided effective communication. The substantial 5.28-point difference favoring online learning in Basic Computer Skills 101 may be attributed to the self-paced nature of the course and the availability of instructional videos. This aligns with research highlighting the benefits of self-paced learning (Tullis and Benjamin, 2011), which allows students to tailor their learning pace to their individual needs and prior knowledge. Further research is needed to explore the interaction between course content, learning modality, and student characteristics to fully understand these varying outcomes. The study offers evidence against catastrophic consequences of rapid online learning adoption, with potential benefits in certain contexts.

This study demonstrates the context-specific effects of online versus in-person learning. A one-size-fits-all approach to online learning is inappropriate. Educational leaders should adopt a nuanced strategy, informed by data and recognizing the diverse impacts of online learning across different subjects and student populations. This includes strategically allocating resources based on the unique needs of each course and exploring opportunities to leverage the benefits of both online and in-person modalities. Future research should investigate the interaction between course design, student characteristics, and learning outcomes to further refine pedagogical approaches.

The study's scope was limited to one large public university in Saudi Arabia. Results may not generalize to other institutions with different resources or student populations. The lack of detailed data on in-class teaching practices and learning activities limits the ability to fully explain the observed variations in student outcomes. The study focused solely on final exam scores; other measures of student learning (e.g., engagement, critical thinking) were not assessed.