The nonlinear model used in relating ln(Y ) to the treatment variables in the combined 1981–1984 data, equation 16.11, can also be fit using linear least squares. If γ is fixed at some value, the...

The nonlinear model used in relating ln(Y ) to the treatment variables in the combined 1981–1984 data, equation 16.11, can also be fit using linear least squares. If γ is fixed at some value, the model is linear in the parameters. Fitting this linear model gives a residual sum of squares that is conditional on the chosen value of γ. Repeating the analysis for a series of values of γ from which the one with the minimum residual sum of squares is chosen will eventually lead to the least squares solution if small enough steps in γ are used. Obtain the least squares solution by this grid search method and compare your results with those obtained from nonlinear least squares. Use γ = 1.0, 1.1, 1.12, 1.13, 1.14, 1.20 as trial values.