Skip to main content
. Author manuscript; available in PMC: 2016 Jan 4.
Published in final edited form as: Q J Econ. 2015;130(3):1329–1367. doi: 10.1093/qje/qjv021

TABLE I.

Effects of the Points and Leaderboard System

(1) (2) (3) (4) (5)
Dependent variable Number of correct answers per day
Post–system change
 dummy
−0.6266***

[0.117]
0.1717**

[0.079]
−0.5391***

[0.117]
−1.2486***

[0.183]
−1.9340***

[0.254]
Mean of dependent
 variable before the
 change
2.57 1.27 2.43 3.47 4.81
Observations 95,342 37,171 22,978 20,427 14,766
R-squared 0.185 0.161 0.159 0.173 0.174
Sample Full Quartile 1 Quartile 2 Quartile 3 Quartile 4

Notes. This table presents OLS regressions of the number of correct answers per day on a dummy on whether the date is after the introduction of the points and leaderboard system. All columns restrict the analysis to the time window between one month before the introduction and one month after it. column (1) presents the results for the entire sample. Columns (2)–(5) present results by quartile of the within-classroom distribution of the total number of correct answers during the month prior to the introduction of the new system. All regressions include student fixed effects. Robust standard errors clustered by classroom in brackets. *** p < .01, ** p < .05, * p < .1.