COVID-19 has revolutionised assessment design and practices in higher education; however, there has been no shift in the objective of enhancing the relationship between assessment and learning that promote the holistic development of students. In this study, we provide an empirical evaluation of the perceived effects of assessment practices (invigilated examination and alternative assessments) on students' mental wellbeing, learning processes, and academic misconduct. A cross-sectional study design was employed for this study in which a self-reported survey instrument was administered to 380 undergraduate and postgraduate students in a public university in Australia. We explored the correlations within defined networks by path analysis via partial least square structural equation modelling (PLS-SEM) framework of SmartPLS 3. Model assessment indexes indicated acceptable convergent, divergent, and construct validity scores for the instrument used. Compared to invigilated exams, students perceived alternative assessments to have significant positive direct effects on stress levels, research skills, learning process, and time management (p < 0.05). In relation to academic misconduct, students generally perceived invigilated exams to restrain such practices, however, the perceived effect was not statistically significant when compared with alternative assessments (p > 0.05). Although, the popularity of alternative assessment practices may have been driven by COVID-19, the pilot findings from this study suggest that these assessment designs and practices have greater potential to promote overall student success and productivity and must be encouraged and utilised in the post-COVID-19 era.