The authors of this rejoinder to Angwin et al.’s (2016) article challenge the conclusions that the COMPAS risk assessment tool is racially biased. They reanalyze the data and argue that Angwin et al.’s findings are flawed due to methodological issues and misinterpretations of the data. The authors emphasize that actuarial risk assessment instruments (ARAIIs) are generally more accurate in predicting recidivism than subjective judgment and that there is substantial research supporting their reliability and validity. They also note that the U.S. criminal justice system is racially disproportionate, and that the use of ARAIs could help reduce this disparity by providing objective decision-making tools. The authors criticize Angwin et al.’s analysis for using a sample of pretrial defendants, which is not the intended population for the COMPAS. They also point out that the COMPAS is designed to predict recidivism for individuals on probation or parole, not for pretrial defendants. Additionally, the authors argue that Angwin et al. misinterpreted the results by collapsing risk categories and failing to properly test for bias according to established standards. They also note that the p-value of 0.0578 was used to conclude that race moderates the relationship between COMPAS scores and recidivism, which is statistically insignificant. The authors conducted their own analysis using the same dataset and found no evidence of racial bias in the COMPAS. They calculated AUC-ROC values and logistic regression models to assess the predictive accuracy and form of the relationship between COMPAS scores and recidivism. They found that the COMPAS predicts recidivism with similar accuracy for both Black and White defendants and that there is no significant difference in the slope or intercept of the relationship between COMPAS scores and recidivism across racial groups. They also found that the interaction term between race and COMPAS score is not significant and adds no predictive power to the models. The authors conclude that the COMPAS is not biased against Black defendants and that the findings of Angwin et al. are based on flawed methodology and misinterpretation of the data. They argue that the use of ARAIs in criminal justice settings is a valid and important tool for making objective decisions and that the concerns about racial bias are not supported by the data. They also emphasize the importance of rigorous testing and proper interpretation of risk assessment tools to ensure their validity and fairness.The authors of this rejoinder to Angwin et al.’s (2016) article challenge the conclusions that the COMPAS risk assessment tool is racially biased. They reanalyze the data and argue that Angwin et al.’s findings are flawed due to methodological issues and misinterpretations of the data. The authors emphasize that actuarial risk assessment instruments (ARAIIs) are generally more accurate in predicting recidivism than subjective judgment and that there is substantial research supporting their reliability and validity. They also note that the U.S. criminal justice system is racially disproportionate, and that the use of ARAIs could help reduce this disparity by providing objective decision-making tools. The authors criticize Angwin et al.’s analysis for using a sample of pretrial defendants, which is not the intended population for the COMPAS. They also point out that the COMPAS is designed to predict recidivism for individuals on probation or parole, not for pretrial defendants. Additionally, the authors argue that Angwin et al. misinterpreted the results by collapsing risk categories and failing to properly test for bias according to established standards. They also note that the p-value of 0.0578 was used to conclude that race moderates the relationship between COMPAS scores and recidivism, which is statistically insignificant. The authors conducted their own analysis using the same dataset and found no evidence of racial bias in the COMPAS. They calculated AUC-ROC values and logistic regression models to assess the predictive accuracy and form of the relationship between COMPAS scores and recidivism. They found that the COMPAS predicts recidivism with similar accuracy for both Black and White defendants and that there is no significant difference in the slope or intercept of the relationship between COMPAS scores and recidivism across racial groups. They also found that the interaction term between race and COMPAS score is not significant and adds no predictive power to the models. The authors conclude that the COMPAS is not biased against Black defendants and that the findings of Angwin et al. are based on flawed methodology and misinterpretation of the data. They argue that the use of ARAIs in criminal justice settings is a valid and important tool for making objective decisions and that the concerns about racial bias are not supported by the data. They also emphasize the importance of rigorous testing and proper interpretation of risk assessment tools to ensure their validity and fairness.