2.5 Residuals <filename>

Once you have fitted the model you should inspect the residual variation around the predicted values. The program calculates the following variables:

p_i

The frequencies of the donor allele predicted by the fitted model.

z_i

Standarized residuals,

$$z_i=\frac{M_i/N_i-p_i}{\sqrt{p_i(1-p_i)/N_i}}.$$

The deviance between observed frequencies and the predicted values divided by the binomial variance around the predicted value under the fitted model.

g_i

The sum of the contribution of all plant varieties k=1 and k=2 to plant i,

$$g_i= \sum_{k=1}^2 g_{ik}.$$

See [1] equation (6). This variable may be a useful predictor of self-fertilization.

These new variables are written to <filename> to the right of the original data variables (see section 2.1). You should import this file into you favorite plotting package and plot the residuals against several of the other variables, e.g. the distance and direction from the donor-field.