The practice of statistics has the power to motivate and support learning across the STEM (Science, Technology, Engineering, and Mathematics) disciplines. When connected with meaningful science contexts, statistical problem solving through the collection of data and subsequent data analysis, supported by contemporary graphing technology, presents opportunities to be innovative and creative in the delivery of learning across STEM curricula. Illustrated in this paper are the statistical outcomes associated with Grade 5 students, who were set the task of solving a mystery about an unknown concentration of a liquid through interpolation of data, which is seen here as finding a missing value in a dataset of two numerical variables where a relationship in the two variables can be represented in a bivariate graph.