1. JABAR H. YOUSIF - Faculty of Computing and IT, Sohar University, PO Box 44, Sohar, PC 311, Oman.
2. EIMAD ABUSHAM - Faculty of Computing and IT, Sohar University, PO Box 44, Sohar, PC 311, Oman.
3. KELVIN JOSEPH BWALYA - Head of Research Development Department, Sohar University, PO Box 44, Sohar, PC 311, Oman.
Recent advancements in artificial intelligence and machine learning have significantly impacted healthcare education by improving efficiency, accuracy, and standardization in patient data analysis. The effects of self-efficacy and test anxiety on academic achievement, using machine learning-based analysis, have attracted attention in many studies, which justify the fact that more research is needed to examine and predicate the real impact of test anxiety on academic achievement. A machine learning method based on the feedforward artificial neural network, the multi-layer perceptrons (MLPs) is used. The study identified five crucial factors for attaining meaningful academic achievement: having a positive mindset, a well- thought-out plan, being accountable for progress, acknowledging potential stress and negative emotions, and monitoring and evaluating one's achievements and efforts. The results showed that having a positive mindset (AR1) was the most important factor for success, with an important rate of 0.997. Monitoring and
evaluating one's achievements (AR5) and a well-thought-out plan (AR2) were also essential factors, with important rates of 0.996 and 0.981, respectively. The study also identified five factors related to test anxiety and academic achievement. The other important factor was AT1 - that the visible signs of nervousness (sweaty palms, shaky Hands, etc.) before a test mainly impacts academic achievement with a rate of .146. Followed by AT7, which stated that some students are more prone to nervousness during exams, ultimately affecting their performance, with an important rate of 0.126. The study also used machine learning to identify distinct patterns in academic resilience and test anxiety factors that affect academic achievement in different student groups. The findings form part of a ‘blueprint’ to inform the development of targeted interventions that cater to the unique needs of student populations and lead to improved academic outcomes. A prediction model has been created to forecast the relevant data and analyze future conditions.
Machine Learning Model, Test Anxiety, Academic Achievement, Multi-Layer Perceptrons, Prediction Model.