We should like to express our thanks to Charles Starnes and Gerald Slatin, who assisted creatively in the statistical analysis upon which this article is based.
Correlational Analysis of Murdocks 1957 Ethnographic Sample1
Article first published online: 28 OCT 2009
1967 American Anthropological Association
Volume 69, Issue 3-4, pages 332–352, June-August 1967
How to Cite
DRIVER, H. E. and SCHUESSLER, K. F. (1967), Correlational Analysis of Murdocks 1957 Ethnographic Sample. American Anthropologist, 69: 332–352. doi: 10.1525/aa.1967.69.3-4.02a00050
- Issue published online: 28 OCT 2009
- Article first published online: 28 OCT 2009
This paper parallels that of Sawyer and LeVine (1966) in some respects: it reduces Murdock's 210 categories to 30 variables, and intercorrelates and factor-analyzes the variables for the six world subdivisions as well as for the entire world. Differences from Sawyer's and LeVine's treatment include the use of Cramér's coefficient to compare all categories, the opposition of the modal category to the others to produce the 30 dichotomous variables, use of the phi coefficient, the testing of the correlations within each of the six areas for significant differences between areas, a different technique for measuring the amount of similarity between areas, a section on factor analysis as a method, and a section contrasting and interpreting correlations in the two areas that differed the most, North America and the Circum-Mediterranean.