In lieu of an abstract, here is a brief excerpt of the content:

A METHOD FOR ANALYZING A CLASS OF RELATED PROBLEMS IN THE PATTERNING OF CENTRAL PLACE ACTIVITIES Thomas L. Bell and Phillip L. Frankland* THE CONTENT ANALYSIS DESCRIBED. Numerous empirical studies exist which attempt to describe the activity patterns of towns of varying sizes in varying spatial environments. (1) Many of these studies have displayed the collected data in the form of an incidence matrix with towns as rows and central place functions as columns. (2) The cells of this dichotomous matrix tell whether a particular central place possessed a particular central function or not. Researchers in different social science disciplines have adopted different research methodologies in order to analyze such matrices. It is the analysis of such data that serves as the point of departure for this paper. There seems to be a fundamen­ tal paradox existing among social science researchers that is perpetuated by strict disciplinary divisions of labor. Although the research endeavors of, say, rural sociologists and geo­ graphers are often the same, each field has developed its own methodol­ ogy for attacking the problem and is often ignorant, or chooses to ignore, the literature of the other field. This paradox certainly comes to light in the case of central place studies. Rural sociologists, from the time of Galpin (3) and Kolb (4) have been interested in the problem of analyz­ ing the functional content of central places. Many of these studies have assumed that the economic impact of the rural community on its sur­ rounding hinterland population has corresponding social impact on those dispersed rural residents. This social impact is manifested by commu­ nity identification, social ties, and contact fields. (5) This thesis of eco­ nomic and social interdependence has not, however, gone without chal­ lenge among rural sociologists. (6) Likewise, geographers from the time of Christaller (7) and Losch (8) have been concerned with the spatial distribution of central places and the effect of this distribution upon the interaction patterns of the surrounding hinterland population. (9) It is at the analysis stage that the paradox discussed above becomes apparent. Both geographers and rural sociologists are concerned with the analysis of incidence matrices of central places and the central func­ tions that they perform, albeit for alternative conceptual reasons. How­ ever, geographers have most often opted for direct factor analysis as * Mr. Bell Is instructor of Geography at the University of Tennessee, Knoxville, and Mr. Frankland is a Ph.D. candidate at the University of Iowa. The paper was accepted for publication in July 1972. V o l. XIII, No. 1 31 the way in which to analyze such matrices, whereas rural sociologists have utilized the technique of scale analysis as developed by Louis Guttman. (10) Although there has recently been explicit mention of the work of the rural sociologists by the geographer Marshall (11), the para­ dox still remains. Most researchers in one discipline seem unaware of the empirical work in the other. ADVANTAGES OF GUTTMAN SCALING AS A METHOD OF ANALYSIS. There are several reasons, made explicit in this section, why geographers should consider Guttman scaling as one of their re­ search tools. The following discussion concerns itself with the Guttman scaling model per se. (12) The Guttman technique was originally designed to rank (or scale) a group of respondents upon the basis of their attitude toward some sub­ ject area. (13) In order for a group of items to be considered “scalable,” they must possess two important properties. First, Guttman scales must be unidimensional. That is, the component items must all measure movement toward or away from the same single underlying object. The substance of the object is relatively unimportant— it may be racial preju­ dice, technological development of societies, or, in the context of this article, ubiquity of central place functions. Secondly, Guttman scales must be cumulative. This is the property which differentiates Guttman scales from almost all other types of scales. Operationally, a cumulative scale implies that the component items can be ordered by degree of difficulty and that respondents who reply positively to a difficult item will always respond positively to less difficult items and vice versa. There are several advantages of Guttman scaling over other tech­ niques that have been...

pdf

Share