Figure 9 - uploaded by Akiva Sanders
Content may be subject to copyright.
Comparing model accuracy for age-based demographic metrics across 540 simulated datasets. The cell values represent the average absolute difference between the specific model's estimate of a metric and the actual value generated by the simulation: lower values imply higher accuracy. Note, the randomly guessing model ("random") and the educated guessing model ("random2") outperform most models.
Context in source publication
Context 1
... Bayesian model outperformed the other models when inferring all age-related estimates and when inferring all simultaneous age-and sex-related estimates (Figure 9 and Figure 10). Importantly, when comparing accuracy scores averaged across the different estimates, the random guessing models outperformed all other models except our Bayesian model and Králík & Novotný's (2003) updated version of Loesch and Czyżewska's (1972) model. ...
Similar publications
Through this research, the aim is to find the mean ridge density in the fingerprints of males and females in the Emirati population to test the three hypotheses discussed above. The research aims at assessing differences in the ridge densities between males and females. It also aims at testing the hypothesis that females have a higher ridge density...
Citations
... We carried out the analysis of all three datasets using new statistical methods that address some of the most intractable issues that have plagued archaeological fingerprint studies in the past and severely hampered their ability to produce accurate results. We describe this new methodology briefly below, but its details, including the meaning of terminology used in this article, instructions for application to other archaeological datasets, and comparisons with other proposed methodologies are included in a separate public access article (Burchill et al., 2023). 6500-720 BCE), in inland western Syria. ...
... The model then generates a new set of guesses based on how reasonable the prior guesses were, and repeats this process millions of times. Again, for a full description of the study's methodology, including the terminology used in this paper, see Burchill et al., 2023. It is important to note that the accompanying methods article also details several lines of evidence that reveal significant problems with the methodology recently proposed by Kent Fowler andcolleagues (2019, 2020), as well as other proposed methodologies, and we strongly discourage their use in the future. ...
In this paper we discuss the application of a new Bayesian statistical model analyze fingerprints on ceramics and other clay objects from the Middle Chalcolithic to Middle Bronze Age periods at the site of Hama, Syria. These results are compared to an analysis of fingerprints on contemporary assemblages at Tell es-Safi, Israel, and Tell Leilan, Syria. These comparisons focus on the Early Bronze Age, a period of rapid social change at all three sites, including settlement nucleation and the development of new systems of social and economic centralization and interdependency. The results of these analyses provide evidence for divergent but overlapping strategies to respond to increases in demand and production capacity during this pivotal period at the different sites under study, including the constraining of gender roles and increased reliance on child labor. In addition, we offer important insights into the process of potter learning, interactions between work and play, and parent-child and child-child relationships that until now could only be guessed.
