a.
Hypothesis test:
H0: βi = 0 ( i= DIFF, PCTHIGH, PCTLOW)
H1: βi ≠ 0
Decision rule:
- Reject H0 + if ᵡ2 > ᵡ2 (0.05, 1) = 3.841
2
- Do not reject H0 if ᵡ < ᵡ2 (0.05, 1) = 3.841
Analysis of Maximum Likelihood Estimates
D Standard Wald
Parameter FEstimate Error Chi-SquarePr > ChiSq
Intercept 1 -3.2041 4.0974 0.6115 0.4342
DIFF 1 0.4150 0.1137 13.3326 0.0003
PCTHIGH 1 3.2578 4.2213 0.5956 0.4403
PCTLOW 1 2.8826 4.2006 0.4709 0.4926
Variable Chi_Squar Compare Decision Conclusion
e
DIFF 13.3326 ᵡ2 > ᵡ2 (0.05,1) = 3.841 Reject H0 Significant
PCTHIGH 0.5956 ᵡ2 < ᵡ2 (0.05, 1) = 3.841 Do not reject H0 No Significant
PCTLOW 0.4709 ᵡ2 < ᵡ2 (0.05, 1) = 3.841 Do not reject H0 No significant
Conclusion at 5% significant level:
+ DIFF is useful in predicting the winners of games in the NCAA basketball tournament.
+ PCTHIGH and PCTLOW are not useful in predicting the winners of games in the NCAA
basketball tournament.
b.
Hypothesis test:
H0: βi = 0 ( i= DIFF, PCTHIGH, PCTLOW, ROUND)
H1: βi ≠ 0
Decision rule:
- Reject H0 + if ᵡ2 > ᵡ2 (0.05, 1) = 3.841
2
- Do not reject H0 if ᵡ < ᵡ2 (0.05, 1) = 3.841
Analysis of Maximum Likelihood Estimates
D Standard Wald
Parameter FEstimate Error Chi-SquarePr > ChiSq
Intercept 1 -4.6750 4.5419 1.0595 0.3033
DIFF 1 0.4705 0.1367 11.8472 0.0006
PCTHIGH 1 5.1695 4.8291 1.1460 0.2844
PCTLOW 1 3.9682 4.4766 0.7857 0.3754
ROUND 1 -0.2727 0.3322 0.6738 0.4117
Variable Chi_Squar Compare Decision Conclusion
e
DIFF 11.8472 ᵡ2 > ᵡ2 (0.05, 1) = 3.841 Reject H0 Significant
PCTHIGH 1.1460 ᵡ2 < ᵡ2 (0.05, 1) = 3.841 Do not reject H0 No Significant
PCTLOW 0.7857 ᵡ2 < ᵡ2 (0.05, 1) = 3.841 Do not reject H0 No significant
ROUND 0.6738 ᵡ2 < ᵡ2 (0.05, 1) = 3.841 Do not reject H0 No significant
Conclusion at 5% significant level:
+ DIFF is useful in predicting the winners of games in the NCAA basketball tournament.
+ PCTHIGH and PCTLOW and ROUND are not useful in predicting the winners of games in
the NCAA basketball tournament.
C.
Home- Court Advantage: A home team will have winning advantages including greater
supports from audiences, and be more familiar with the court. These advantages will have
positive psychology impacts on the home team to win the game.
Personal income of players in the team: in reality the wealthier city will offer higher
personal incomes for players. The higher income will bring the higher motivation for the
team to put more efforts to win the game.
The percentage of the higher seed number won the lower seed number won in the
past: The higher percentage the higher seed number won the lower seed number in the
past, the higher possibility that the higher seed number will win the lower seed number in
this game.
