<p>Evaluating researcher impact is a critical aspect of scientometrics, influencing key academic decisions such as promotions, funding allocations, and prestigious awards. While traditional bibliometric indicators such as the h-index and its various extensions are widely adopted, ongoing debates persist regarding their accuracy in capturing true research impact. This study introduces a novel index aimed at improving researcher evaluation in the field of computer science. Using a dataset of 1200 researchers, evenly split between 600 award recipients and 600 non-recipients, we systematically evaluate individual bibliometric indices to identify the most effective parameters. These top-performing parameters are then combined using advanced statistical models including the geometric mean, harmonic mean, and contra-harmonic mean, etc., to determine the optimal aggregation method for impact assessment. The outcome of this study indicate that the h2-upper index and the AWCR index demonstrate the highest predictive accuracy among individual indicators. Moreover, the contra-harmonic mean proves to be the most effective statistical model for combining parameters. Based on these insights, we propose a new index derived by applying the contra-harmonic mean to the best-performing parameter pair. This enhanced framework offers a more robust and reliable approach to researcher evaluation, supporting improved academic assessment and recognition in computer science.</p>