[TIXAUC.sub.j] = [A.sub.0] + [B.sub.1] [SPREAD.sub.j] + [B.sub.2] [INCOME.sub.j] + [B.sub.3] [POP.sub.j] + [B.sub.4] GDAY + [B.sub.5] ROUND + [B.sub.6] TIXFACE + [B.sub.7] [CURWIN.sub.j] + [B.sub.8] [PREVWIN.sub.j] + [B.sub.9] DIV + [B.sub.10] BIDS + [B.sub.11] DAYSBEF + [B.sub.12] CERTAIN + [B.sub.13] FORMAT + [B.sub.14] PFB + [B.sub.15] NFB + [B.sub.16] LENGTH + [B.sub.17] TRANS
As displayed in Table 3, the variables included in the selected model were population (POP) (p = .008), total number of transactions per game (TRANS) (p < .01), face value of the ticket (TIXFACE) (p < .01), winning percentage from the previous year (PREVWIN) (p < .01), day of the game (GDAY) (p < .01), number of days before the game (DAYSBEF) (p <.01), per capita income (INCOME) (p <.01), playoff round (ROUND) (p < .01), total number of bids (BIDS) (p < .01), and current winning percentage (CURWIN) (p < .01.
Variable Correlations with Final Ticket Price at Auction Variable Secondary Ticket Value TRANS .014 INCOME .338 * POP -.443 * TIXFACE .341 * CURWIN .273 * PREVWIN .224 * SPREAD .107 * BIDS .304 * DAYS .108 * PFB .025 NFB .099 * LENGTH -.123 * GDAY .083 * ROUND .563 * CERTAIN .076 FORMAT .112 * DIV .164 * Note: * Correlation is significant at the .05 level.