Electronic Library of Scientific Literature - © Academic Electronic Press
Volume 50, 2002, No. 1
This study deals with various possibilities of application of statistical – quantitative methods in anthropology (ethnology), specially in the stage of ethnographical description (analysis of empirical data). It refers to the shortcomings in methodological education and praxis of Slovak ethnography where we can state that quantitative methods are taught and used only very seldom. We claim that almost all of the ethnographically studied phenomena have two levels of understanding – the individual and collective level. It means, we can study it also using concepts such as frequency and distribution in population of individuals. Quantitative methods are irreplaceable when our generalisations refer to collective level of social and cultural phenomena. In statistical approach, an individual phenomenon is considered to be only one part of a collective phenomenon. Our knowledge of an individual phenomenon we can not simply generalise without the risk of errors, but ethnographers must generalise during the process analysis and interpretation of empirical data. Statistical methods provide strong tools for finding solutions of this task.
For better understanding of further comments we introduce fundamental concepts of mathematical statistics such as statistical set, statistical unit and statistical feature (also variable). In social sciences, we analyse 3 kinds of data (variables) – nominal, ordinal and cardinal. These different kinds of data need also different methods of analysis. We stress that inadequate choosing of method leads to meaningless results.
There are three basic groups of statistical methods applicable in ethnographical data analysis. First group known as univariate statistics is oriented mainly to description of one variable (calculating of central value and measure of variability and describing of frequency distribution). That is why it is often called descriptive statistics. Other interesting part of univariate statistics is testing of statistical hypotheses. When we try to find relations between two variables defined on the same statistical set, we can use methods such as regression (finding the best fit) and correlation (calculating the measure of formal interdependence between variables). These methods belong to the second group of statistical methods known as bivariate statistics. Results of both two groups of statistical tools are used in the third group of statistical methods studying relations (formal interdependences) of more than two variables. To this third group known as multivariate statistics belong modern quantitative methods such as factor analysis and multidimensional scaling, which could be a very strong tool for analysis of complex sociocultural systems, both behavioral and mental (cognitive).
All of methods listed in the study are suitable for ethnographical data analysis under some conditions requiring better understanding of mathematical statistics. That is why this study contains more detailed introduction to principles of basic methods and the logic of different data kinds, but it is not enough – there is a significant lack of textbook of quantitative methods for ethnography in Slovak or Czech ethnological literature. In the end of study is listed commonly used statistical software suitable for social sciences with short comments.
SN 1/2002, p. 5-24
Two major lines can be traced in the article. The first one deals with a critical view on population census as a source in the area of ethnological and social-historical research, while the second concentrates on the use of this source in the family structure research according to the theory of ecotypes.
The official count of inhabitants of a certain region or country or the so-called census belongs to the group of quantitative sources. It provides an insight into the structure of the population in a special time instant. With other words it reflects a static picture of the society on the contrary to different sorts of registers like for example official lists of births, marriages and deaths. These lists perceive and record the changes of population in the course of time. As a source of information the census offers a whole string of benefits, for instance the opportunity of using statistical quantifying methods during the evaluation procedure. At the same time it allows the concentration on the individual cases.
There exist more possibilities how to deal with the information gained from census sheets, the so called aggregative method rests in work on several demographic phenomena, the next approach resides in processing individual sheets and results in the formation of different graphs of family structure and the third method produces genealogical tables.
The theoretical base of this article is the concept of ecotypes developed by Scandinavian cultural anthropologists and Austrian social historians. It is based on the premise that the local dominance of different modes of production is dependent on the exploitation of specific natural resources of the physical environment. This approach is enriched by the results originating from fieldwork in the Nord-western Slovak region Kysuce, Nová Bystrica village. The combination of various sorts of written and oral sources (census from 1930, marriage, birth and death registers, village chronicle and interviews with local inhabitants) allowed to show a quite exact family structure in the first half of 20th century and its relationship to labour organisation in agricultural society connected with the natural environment. The research showed that in the two major occupations of the inhabitants of Nová Bystrica – peasantry and cattle raising, the necessity of labour forces was satisfied in prevailing cases with the family members. If this solution was not possible, the family dicided to hire fixed stock of permanent workers in cattle raising meanwhile in the case of the peasantry the missing labour forces were hired short-term during the period of summer work-harvest.
The aim of the work was also to present how fruitful the mentioned combination of qualitative and quantitative sources, showing the structure of the family and relationships existing between individual members, can be.
SN 1/2002, p. 25-44