Baxendell

The Baxendell method is a statistical technique used in archaeology, particularly in pottery analysis, to address issues arising from the "size effect" in compositional data. The size effect refers to the spurious correlations that can arise in compositional data (data where variables represent proportions of a whole), where the sum of all variables is constrained to a constant (e.g., 1 or 100%).

The core problem is that if one component in a composition increases, one or more other components must decrease, leading to negative correlations even if no genuine relationship exists between the components. The Baxendell method is one approach aimed at correcting for these inherent dependencies.

Specifically, the Baxendell method typically involves calculating a series of ratios between the elements or components being analyzed. By transforming the original elemental concentrations into ratios, the method aims to remove or minimize the constraints imposed by the closure effect (the constant sum constraint). This allows for more meaningful statistical comparisons and interpretations.

The exact implementation of the Baxendell method can vary depending on the specific research question and the nature of the data. Common ratios include elemental ratios (e.g., the ratio of element A to element B) and ratios of each element to the total concentration of a select group of elements.

While the Baxendell method can be a useful tool, it's important to understand its limitations. The choice of which ratios to calculate can significantly influence the results, and the method does not eliminate all potential issues related to compositional data. Other methods, such as centered log-ratio (clr) and isometric log-ratio (ilr) transformations, are also frequently used to analyze compositional data and may be more appropriate in some cases. Therefore, the application of the Baxendell method, like any statistical technique, should be carefully considered in light of the specific research context.