Accounting for multiple effects and the problem of small sample sizes in osteology: a case study focussing on entheseal changes

“

The method used to record entheseal changes has previously been shown to identify different occupational categories and is therefore being applied to study occupation in archaeological contexts. However, in this sample of archaeological (but identified) skeletons occupation this effect was not found. Odds ratios showed differences between occupational groups, but generalised linear methods showed that the effect was predominantly one of age rather than occupation. This is important because in most archaeological samples age-at-death is difficult to pinpoint and difference which may be attributed to occupation could be an age-effect. This paper demonstrates the problems of testing for small effects (i.e. the effect of biomechanics on the development of entheseal changes) in a typically small sample of identified skeletons from the UK. Generalised linear models, as used here, are a really robust method for biological analysis because of their ability to take into account multiple factors. This makes them ideal for bioarchaeological research, but the small sample sizes typically involved make such applications hard. In this study four sites with identified skeletons from the UK were pooled to create a reasonable sample size, but there is a clear age bias towards older individuals, as well as lots of missing data. The generalised linear methods did show the effect of age for many of the entheseal changes, body size was also important for some variables, but only one showed an effect of activity.

”