In this paper an elementary exploration is given of some of the features of Q-discrimination analysis, an adaptation by Johnson of Atkin's Q-analysis for use in clustering work. The method is summarised. It is shown how the performance of the method varies considerably, depending upon whether the data involved are integer or noninteger, over a given range of values, and on whether they are ordinal or interval. An ambiguity over the criteria for determining discrimination levels is also considered.
Get full access to this article
View all access options for this article.
References
1.
AtkinR H, 1974a“An approach to structure in architectural and urban design. 1. Introduction and mathematical theory”Environment and Planning B151–67
2.
AtkinR H, 1974b“An approach to structure in architectural and urban design. 2. Algebraic representation and local structure”Environment and Planning B1173–191
3.
AtkinR H, 1974cMathematical Structure in Human Affairs (Heinemann Educational Books, London)
4.
AtkinR H, 1975“An approach to structure in architectural and urban design. 3. Illustrative examples”Environment and Planning B221–57
5.
AtkinR H, 1977Combinatorial Connectivities in Social Systems (Birkhäuser, Basel)
6.
AtkinR H, 1981Multidimensional Man (Penguin Books, Harmondsworth, Middx)
7.
BeaumontJ RGatrellA C, 1982CATMOG 34: An Introduction to Q-AnalysisGeo-Abstracts, Norwich, England
8.
ChapmanG P, 1981“Q-analysis” in Quantitative Geography: A British View Eds WrigleyNBennettR J (Routledge and Kegan Paul, Henley-on-Thames, Oxon) pp 235–247
9.
DuranB SOdellP L, 1974Cluster Analysis: A Survey (Springer, New York)
