Reply to Thread

Message:: [QUOTE="Roerbakmix, post: 4462838, member: 100731"]Interesting breakdown. You did make one error however in your calculation: the average of the sale price is not a weighted average (i.e. the proportion of the subgroups differ). Here is the correct grand total:[ATTACH=full]1111948[/ATTACH](the absolute number does not count up to 800 because of rounding in the percentages) Also, note that your Bryzantine and 'other' is not included in this, so the grand total would be higher. *Edit. Under the assumption that the distribution of grades ("very low grade cull", "low grade" etc.) only applies to Roman Imperial, this would be the correct grand total:[ATTACH=full]1111951[/ATTACH]*Provincial/other: Greek: 1.1%, Roman Provincial: 3.4%, Seleucid: 0.7%, Celtic/other: 1.3%, Roman Republican: 0.23% The weighted average would be $1,116.27/800 ~ $1.40. Finally, it would be interesting to have the sale price per individual coin. Hypothetically, your average sale prices may include a few extreme outliers (suppose someone without knowledge buys a "very low grade cull" for $300 ) - unlikely, but this will of course make the average price difficult to appreciate. A standard error of the mean makes interpretation a bit more easier, and the results somewhat more interpretable.[/QUOTE]