chapter  2
Simple Models
Definitions of Error and Parameter Estimates
ByCharles M. Judd, Gary H. McClelland, Carey S. Ryan
Pages 15

In this chapter, the authors’ consider the very simplest models—models with one or even no parameters. The simple model provides a useful first-cut description of data. Measures of location or measures of central tendency are the traditional names for the parameter estimates of ß0 developed from the different definitions of error. When the authors' use the sum of absolute errors as the aggregate index of error, it is customary to use the median absolute error or deviation from the prediction to represent the typical error. A useful index of error is the remaining error per remaining potential parameter. The sum of squared errors is the formal equivalent of adding up the error squares. Besides removing the signs, the squaring has the additional effect of making large errors more important. The sum of absolute errors is the formal equivalent of summing the line lengths.