Table 6.

Peer Effects with Alternative Measures of Peer Ability

	(1)	(2)	(3)	(4)	(5)
Measure of Peer Ability:	Average Ability	Max Ability	Min Ability	Tiger Woods is partner	Avg. Ability × (Avg. Ability − Own Ability)
Own Ability	0.673 (0.039)	0.673 (0.039)	0.673 (0.039)	0.675 (0.039)	0.673 (0.039)
Peer Ability	−0.035 (0.040)	−0.021 (0.032)	−0.032 (0.040)	−0.346 (0.464)	−0.016 (0.014)
R2	0.152	0.152	0.152	0.152	0.152
N	17492	17492	17492	17492	17492

	(6)	(7)	(8)	(9)
Measure of Peer Ability:	1{any partner in top 10%}	1{any partner in top 25%}	1{any partner in bot 25%}	1{any partner in bot 10%}
Own Ability	0.674 (0.039)	0.672 (0.039)	0.674 (0.039)	0.674 (0.039)
Peer Ability	0.027 (0.070)	0.050 (0.059)	−0.039 (0.069)	−0.175 (0.141)
R2	0.152	0.152	0.152	0.152
N	17492	17492	17492	17492

Notes:

^aColumn (1) is reproduced from Table 3. Other columns present results from modifying baseline specifications as specified in equation (3) to support heterogeneous peer effects.

^bThe dependent variable is the golf score for the round.

^cThe Ability variable is measured using the player’s handicap.

^dStandard errors are in parentheses and are clustered by playing group.

^eAll regressions weight each observation by the inverse of the sample variance of estimated ability of each player.

^fAll regressions include tournament-by-category fixed effects.

^gIn column (9), average ability of playing partners is also included in regression; the estimated coefficient for this variable is −0.014 (0.048).