Figure 4.19 Correlations between alternative hypotheses linking city size with spacing. The pecked line denotes equal correlation values for the two hypotheses. Sources: Thomas. 1962, p. 27; Gibbs, 1961, p. 45S.
Gibbs (1961, pp. 451-9) has confirmed Thomas's hypothesis for the spacing of the major cities in six countries (Brazil, Canada, France. Italy, Mexico, and the Netherlands} As Figure 4.19a indicates, a higher correlation coefficient was found between the size of metropolitan area and the distance to the nearest larger metropolitan area, than between size and the distance to the nearest metropolitan area of the same type. Despite variations in the values of the coefficient, the consistently higher values for the second relationship are striking.
•
(c) Spacing as a multivariatefunction. In one of the few studies in which spacing has been treated in a wider context, King (1961) carried out a multiple regression analysis of several alternative hypotheses. Spacing was seen as a function of the size of a town, its occupational structure, and the characteristics of the zone in which it was located. A sample of 200 towns was drawn at random from the 1950 United States census giving a range in town size from
Figure 4.19 Correlations between alternative hypotheses linking city size with spacing. The pecked line denotes equal correlation values for the two hypotheses. Sources: Thomas. 1962, p. 27; Gibbs, 1961, p. 45S.Gibbs (1961, pp. 451-9) has confirmed Thomas's hypothesis for the spacing of the major cities in six countries (Brazil, Canada, France. Italy, Mexico, and the Netherlands} As Figure 4.19a indicates, a higher correlation coefficient was found between the size of metropolitan area and the distance to the nearest larger metropolitan area, than between size and the distance to the nearest metropolitan area of the same type. Despite variations in the values of the coefficient, the consistently higher values for the second relationship are striking.•(c) Spacing as a multivariatefunction. In one of the few studies in which spacing has been treated in a wider context, King (1961) carried out a multiple regression analysis of several alternative hypotheses. Spacing was seen as a function of the size of a town, its occupational structure, and the characteristics of the zone in which it was located. A sample of 200 towns was drawn at random from the 1950 United States census giving a range in town size from
การแปล กรุณารอสักครู่..
