Adams et al. (2018) point out that outliers might be driving the negative relationship between fund size and performance in chen et al. (2004). These outliers are due to style misclassifications in the 2004 CRSP Mutual Fund Database. They propose robust regressions to remove outliers. We point out that the ideal way to address this issue is simply to clean up these style misclassifications as recent papers in the literature have done. Removing outliers can skew inference if small (large) funds have positively (negatively) skewed returns due to diseconomies of scale. We show that after cleaning up style misclassifications, the negative relationship between fund size and performance remains robust regardless of estimation strategies.