[Archive]

Posted By: Glenn

This relates to my area of expertise, so I feel an obligation to explain it. The proposed hypothesis is that the average 150 series card is in lower grade than the average 350 (or 350 or 460) series card. You won't be able to test the validity of your hypothesis by surveying individual collectors unless you can assume (and clearly you cannot) that the collectors do not factor the condition of a card into their purchasing decision AND that condition is unrelated to the value of the card. What you need to do is something called a chi-square analysis, and the data that are required as input have to be a representative sample of the total populations of the cards in each condition. The fact that a card's condition correlates with its probability of having been submitted for grading is inconsequential as this should be equally true of cards from 150 and 350 and 460 series. Ideally, we would have the complete population reports by series from all the major graders. The broader range of grades we have the data from, the more Power (technical term) we'll have in conducting our analysis (i.e., the easier it will be for us to determine that our hypothesis is correct, if in fact it is). So far, we have the data Ted provided of the reports from SGC across 3 grades. Give me a few minutes so I can perform the appropriate analysis.

Glenn