I recently got some Uranium–Thorium dates back from a lab. They are reported with a ±2σ uncertainty. Several of the measurements are replicates from the same specimen. That is, a few pieces from a single rock (a single annual layer from a speleothem, by definition deposited during the same year) were analysed on the same instrument.
My understanding of statistics is quite limited, but I know mean of the measurements should approach the true age (assuming no systematic error) as the number of measurements increases. If I want to summarise my results, what is the appropriate way to treat the uncertainty? Or perhaps this is not the best way to think of these results.
| Age (ka) | ±2s |
|---|---|
| 11.7 | 0.85 |
| 12.4 | 0.39 |
| 11.2 | 0.76 |
| mean: 11.8 | ?? |
Please enlighten me!
