Statistical analysis in archaeology
I have always been interested in statistical methods, but my interest strongly increased when in 1982 I joined a course in multivariate statistics at the University of Tromsų. The course not only introduced me to Correspondence Analysis, which at that time was rather unknown, but also to its computation. I carried the code, written in a statistical programming language called Genstat, with me from Tromsų and had it installed on the mainframe computer at Aarhus University (1984, 1985). This installation provided the basis for the early use of Correspondence Analysis in Danish Archaeology as reflected in the book Multivariate Archaeology (1988).
In 1990, I implemented Correspondence Analysis and Principal Component Analysis on DOS PC's through a program called KVARK written in Pascal. This was frequently used in Denmark, but was discontinued following the spread of Irwin Scollars successful program Basp and its successor WinBasp.
In 2005, I rewrote the program into an add-in to Microsoft Excel called CAPCA. Correspondence Analysis, Principal Components Analysis and Metric Scaling are here fully integrated with Excel worksheets and the graphical presentations occur through Excel graphs.