Extended Data Figure 10: Results of pIR-IRSL analyses at ATM.
From: Early Middle Palaeolithic culture in India around 385–172 ka reframes Out of Africa models
The panels show a–d separately for ATSP-1, ATSP-3, ATSP-5, ATSP-7, ATSP-9, ATSP-11 and ATSP-12. a, Typical feldspar shine-down curve. Open circles, natural irradiation; black circles, after beta irradiation. b, Typical growth curve. c, Radial plot representing scatter in estimated palaeo-doses. d, Typical g-value data. In c, the x axis represents precision (1/σ) and the y axis represents the standardized estimate (Z/σ) with respect to the mean value (here, Z0); the circular scale represents estimate values of Z (or dose) in Gy. Points with a larger x value have higher precision. The grey band shows the 2σ envelope of the calculated mean. In d, the probability value is the probability of the χ2 distribution and the dashed curves represents the 2σ hyperbolic error envelope, defining the confidence region around the best-fit line. On average, the probability value should be equal to 50%. Larger values indicate that the overall scatter of the data (around the mean regression line) is smaller than the uncertainties of each data point. A smaller probability demonstrates that the scatter is larger than the uncertainties of each data point, and the mean square weighted deviation (MSWD) measures the goodness of fit of the model to the data. MSWD = 1 when the data fit a univariate normal distribution as a function of t or log10(t). MSWD > 1 when the observed scatter exceeds predicted analytical uncertainty (data are over-dispersed). MSWD < 1 when the observed scatter is less than the predicted analytical uncertainty (data are under-dispersed).