7. Regression analysis

Once the significant influence of a factor (or interaction) has been demonstrated in the case of systematic selection of the levels of the controlled factor, it is natural to look for a function representing the phenomenon. The most frequent case is that of a monotonic function (increasing or decreasing), which can always be – by means of an anamorphosis, or – represented as a linear function. For example, the parabola y = ax ² becomes a straight line y = at by setting , the exponential y = ae ^+bx becomes a straight line t = – b x + m by setting t = ln a and m = ln y, and so on....