Regression correction

[From Rick Marken (950124.1255)]

Rick Marken (950124.0845) said:

The intercept of the regression equation gives an estimate of r only under
(at least) the following special circumstances:

1. The average of the predictor (d) must be 0.

2. The dependent variable, h, must be measured in units of the controlled
variable, c.

After writing this I remembered that h must not be stored in screen units
because maxx/2 is added to c before the cursor is displayed. So h is stored
0 centered. d is also stored 0 centered but, in his regressions, Bruce
coverted d to screen units by adding maxx/2 to d. The intercept of the
regression was equal to the reference value for the controlled variable (319)
because, when is d = 319 h is about 0. Since h = r - kd and k is negative,
the intercept is 319 when d is 319 to predict an h of 0. So the intercept of
the regression equation gives an estimate of the reference state of the
controlled variable when the 0 of the dependent variable (h) corresponds to
screen center and the predictor variable, d, is measured in units of the
controlled variable.

There's got to be an easier way to not make a living;-)

Best

Rick