Characteristics of Latin American Countries

Abstract

The authors have made a statistical construction of the "average" Latin American country. Factor analysis produced three central indices—Size, Developmental Level, and Exports to the United States—which account for considerable differences between nations. Similarities in certain of the 16 basic factors, such as literacy, are also pointed out. Paul Deutschmann, who died unexpectedly last year, was Professor at Michigan State University, and one of the finest and most cordially liked younger scholars in the field of journalism and communications research. Professor McNelly is at Michigan State, following an extended period of research in Latin America.S statistics of national states increas-A ingly become available it has been observed that there are a number of significant relationships between some of them. For example, Lerner, using UNESCO data, showed, that literacy and percent of urbanization were highly related around the world, and that media development also related to these two concepts. McClelland also has shown that a measure such as electrical power consumpiton is highly related to national income and other development indices. A more recent study is that of Gibbs and Martin.'The present investigation was designed to examine a number of such national indices for 20 Latin American countries, to determine the degree of intercorrelation among them, and to determine by an approximation to factor analysis how many different aspects of these nations they tell us about. In addition, we will analyze the pattern of interrelations for particular variables of special interest. Such analysis is dependent upon the quality of statistics used,2 but not to as great an extent as we might expect. Interrelationships which are very strong will emerge even if there are fairly substantial &dquo;errors&dquo; in the figures for several nations. To the extent that there are random errors, they can only serve to reduce relationships. Systematic errors, such as could arise from the utilization of some indices to estimate other key figures, can on the other hand introduce correlations which might be spurious or artificial. These two forces, of course, are working in a countervailing fashion-the one to increase, and the other to lower the estimates of the &dquo;true&dquo; relationships between variables.We have chosen to work with the 20 They tend to have a common historical and language background, with the exception of Brazil and Haiti. All are independent national states.Since we are working with a relatively homogeneous region in terms of backgrounds, we may well expect stronger relationships than have been developed in other similar studies which have utilized states from various part of the world with quite divergent cultures. On the other hand, despite the homogeneity of background for the chosen set, there is a considerable range in most of the indices in which we are interested, another factor necessary to produce some magnitude in correlation.In Table 1 we present the 16 indices which were used and three m...

A Traditional Community in the Chao Phraya River Basin: Classification and Characteristics of a Waterfront Community Complex

“…However, it is necessary to exercise caution on these points, since on the basis of the correlation coefficient alone, no matter how large, it cannot be said that one variable creates another. In order to do that, logic, additional evidence, and/or analysis must be applied as the measure states the degree of the relationship and nothing more (Deutschmann & McNelly, 1964). The interpretation of the PCI with respect to the values is described in the next section.…”