<p>Shear wave velocity (<i>V</i>_<i>s</i>) is vital in geotechnical engineering, indicating soil behaviour under dynamic forces like earthquakes. Calibration of the relationship between <i>V</i>_<i>s</i> and Standard Penetration Test (SPT) blow count (<i>N</i>) in sandy soil bridges the gap between an easily obtainable in-situ test and a crucial dynamic soil property. This research hypothesizes that the natural logarithms of <i>Vs</i> (i.e. <i>ln (Vs)</i>) and <i>N</i> (i.e. <i>ln (N)</i>) exhibit a consistent proportional relationship, with a slope that remains uniform and the intercept varies according to the geographic origin due to data uncertainties. Based on the analysis, it was observed that the variation is predominantly in the intercepts rather than the slopes, as indicated by the literature reviews. A two-level Bayesian multilevel regression model was employed, with varying intercepts for regional differences. The first level captures within-region variation, where each region has a unique intercept. The second level captures between-region variation, assuming a consistent slope across regions. This structure allows for accurate modeling of shear wave velocity and SPT blow count relationships, accommodating regional differences effectively. Bayesian multilevel model acknowledges data clustering, improving predictions by using information from data-rich regions to inform data-scarce areas. This two-level structure enhances understanding of regional variability and model robustness. The Bayesian framework further quantifies uncertainty. This validated approach offers a practical method for estimating <i>V</i>_<i>s</i> in sandy soils, streamlining site investigations, and facilitating the development of regional <i>V</i>_<i>s</i> maps for better seismic hazard assessments. This research provides a robust, broadly applicable model for geotechnical engineering.</p>