This session from Yale's Introduction to Medical Software course walks through how to build a validation plan for a new piece of medical software. The instructor explains how to turn a vague user need into a specific, measurable, testable goal, then breaks a validation plan into its four working parts: the goal, the rationale behind it, the methods used to test it, and the plan for analyzing the resulting data. The lecture stresses bringing in a statistician early, before data collection begins, so the study design can actually support the statistical analysis it will need later. Examples compare a new software solution against an existing method or the current standard of care, showing how those comparisons get built into the plan. It is a short, practical module aimed at students who already have the course's earlier context on regulatory and design basics.