A new approach to radiocarbon summarisation: Rigorous identification of variations/changepoints in the occurrence rate of radiocarbon samples using a Poisson process
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
A commonly-used paradigm to estimate changes in the frequency of past events or the size of populations is to consider the occurrence rate of archaeological/environmental samples found at a site over time. The reliability of such a “ dates-as-data ” approach is highly dependent upon how the occurrence rates are estimated from the underlying samples, particularly when calendar age information for the samples is obtained from radiocarbon ( 14 C). The most frequently used “ 14 C-dates-as-data ” approach of creating Summed Probability Distributions (SPDs) is not statistically valid, or coherent, and can provide highly misleading inference. Here, we provide an alternative method with a rigorous statistical underpinning that also provides valuable additional information on potential changepoints in the rate of events. Furthermore, unlike current SPD alternatives, our summarisation approach does not restrict users to pre-specified, rigid, summary formats (e.g., exponential or logistic growth) but instead flexibly adapts to the dates themselves. Our methodology ensures more reliable “ 14 C-dates-as-data ” analyses, allowing us to better assess and identify potential signals present. We model the occurrence of events, each assumed to leave a radiocarbon sample in the archaeological/environmental record, as an inhomogeneous Poisson process. The varying rate of samples over time is then estimated within a fully-Bayesian framework using reversible-jump Markov Chain Monte Carlo (RJ-MCMC). Given a set of radiocarbon samples, we reconstruct how their occurrence rate varies over calendar time and identify if that rate contains statistically-significant changes, i.e., specific times at which the rate of events abruptly changes. We illustrate our method with both a simulation study and a practical example concerning late-Pleistocene megafaunal population changes in Alaska and Yukon. • Summed probability distributions (SPDs) do not provide a valid, or coherent, approach to summarise sets of 14 C dates. • We introduce a statistically-rigorous, fully-Bayesian, alternative that ensures more reliable 14 C-dates-as-data analysis. • Information on the varying occurrence rate of archaeological/environmental 14 C samples over calendar time is provided. • The calendar timings of any substantial changes in the sample occurrence rate are identified and estimated. • Code and a user guide are available in the carbondate R library on CRAN and at https://tjheaton.github.io/carbondate/ .
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.003 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.002 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it