Methods of correlation and regression analysis, linear and curvilinear

Thirty years have elapsed since the original edition of this book was written—years of political tensions and upheaval ^nd of enormous progress in technical development. This last has been reflected in changes in some of the examples cited—from an automobile with two-wheel brakes in the 1920’s to the orbit of an earth satellite in the late 1950’s, and from methods for using hand calculators and card tabulators to those for electronic computers. Despite this technical progress, the basic elements of correlation analysis continue unchanged. The major emphasis, however, has shifted from correlation to regression, and the wide range of uses of the method in varied fields has led to many specialized applications or modifications. This is especially true in econometrics. Here the long controversy over mutually intercorrelated variables has finally produced an effective simultaneous- equation method for dealing with situations where single-equation solutions are inadequate; but apparently such situations are relatively infrequent.