Stress testing involves running a numerical model, constructed from specific data about your proposed project, and subjecting your construct to a variety of equations that abstractly test the durability of your design. These tests are especially important for testing building materials and contraptions people will be physically inside of; examples include planes, trains and cars. Most of these equations are impossibly complex to perform by hand, so engineers utilize computer models and programs specifically designed to run stress testing. These programs can be generalized, or specialize in stress-testing specific components of the object in question.
Computer Aided Design (CAD) is a computational innovation which allows engineers to take the specifics of their engineering design and construct their project in a virtual environment. This program utilizes sophisticated equations to model real-world conditions, and how the engineer's design would behave if actually constructed. This program can test items such as the effect of forces on their design, how the design would behave in motion tests and give data on the physical properties of the experimental design. This tool is used both by professional engineers in their development stages, and by students to learn the principles of engineering.
Regression analysis is a core tool of engineers and scientists of all stripes to test causality with statistical certainty. Regression analysis is more advanced than a simple correlation analysis, which simply tests for a quantifiable relationship in two sets of data. Correlation analysis only tests for the existence of a relationship, not for how one set of data might have a causal effect on the other set. A regression analysis does just this: tests to see if one set of data has a statistically significant effect on the other set of data. Engineers take the data from experiments they perform on their devices to see if they can statistically prove what they created works.