a procedure for analyzing the relationship between an independent variable (x) and a dependent variable (y) where the regression equation involves exponential forms of x. That is, the changes in y are not consistent for unit changes in the x variable(s) but are a function of the particular values of x. For example, a nonlinear regression model is given by y = a + b1x1 + b2x12 + e, where a indicates the place where the line of best fit crosses the y-axis, b1 indicates the number of units that y changes when x is changed by 1 point, and e is error. Also called curvilinear regression. Compare linear regression. See regression analysis.