The current research evaluation system is inherently contradictory to the principles of open science. Existing evaluation frameworks remain heavily reliant on citation-based metrics, which often fail to fully recognize the broader value generated by preprints, social dissemination, and open data sharing. This study employs an exploratory case study approach to examine the preprint impact and societal impact of scholarly outputs among astrophysics researchers and to analyze the usage statistics of open data platforms among scholars in the earth sciences. The findings reveal that traditional research evaluation systems significantly underestimate the early academic impact of preprints, inadequately capture scholars' societal influence, and fail to effectively recognize the value created by open data infrastructure. To address these limitations, this paper proposes an open science-oriented scholar evaluation framework that integrates academic impact, societal impact, open data/infrastructure, and open science practice into a unified assessment system. Through scholarly case studies, this paper provides empirical evidence and offers practical insights for reforming research evaluation systems in the era of open science.