A machine learning-based analysis for the effectiveness of online teaching and learning in Pakistan during COVID-19 lockdown

Abstract

Background

The COVID-19 pandemic has significantly disrupted daily life and education, prompting institutions to adopt online teaching.

Objective

This study delves into the effectiveness of these methods during the lockdown in Pakistan, employing machine learning techniques for data analysis.

Methods

A cross-sectional online survey was conducted with 300 respondents using a semi-structured questionnaire to assess perceptions of online education. Artificial intelligence methods analyzed the specificity, sensitivity, accuracy, and precision of the collected data.

Results

Among participants, 42.3% expressed satisfaction with online learning, while 49.3% preferred using Zoom. Convenience was noted with 72% favoring classes between 8 AM and 12 PM. The survey revealed 87.33% felt placement activities were negatively impacted, and 85% reported effects on individual growth. Additionally, 90.33% stated that online learning disrupted their routines, with 84.66% citing adverse effects on physical health. The Decision Tree classifier achieved the highest accuracy at 86%. Overall, preferences leaned toward traditional in-person teaching despite satisfaction with online methods.

Conclusions

The study highlights the significant challenges in transitioning to online education, emphasizing disruptions to daily routines and overall well-being. Notably, age and gender did not significantly influence perceptions of growth or health. Finally, collaborative efforts among educators, policymakers, and stakeholders are crucial for ensuring equitable access to quality education in future crises.