Using linear and nonlinear models to design experiments, regress coefficients, and infer confidence intervals for both the regressed coefficients and the predictions for systems with errors in one or several variables.