Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Using data of twin students graduating from junior high schools between 2002 and 2004 in Taiwan, we estimate the effects of peers on high school joint entrance examination performance. To alleviate the potential endogeneity of peer educational achievement, linear models with twin fixed effect and instrumental variables are estimated. Quantile, quantile with twin fixed effect and quantile with instrument variables regressions are also implemented to determine whether estimated peer effects differ at various locations of the testing scores' conditional distributions. Positive and statistically significant peer effects are found to exist at the mean and at different quantile levels.