Software Predicts Fate Of Death Row Inmates

Chicago (IL) - This one may give you chills. Scientist claim to have found a way to predict whether a death row inmate will be executed or not with an accuracy of 90-percent. They found a strong relation between certain profiles of an inmate and the likelihood of execution. But it may not be the profile you are thinking of.

You could call it a cold calculation. But you can’t deny that there are always has been prejudice and there always have been suspicions and conspiracy theories about how death row inmates are selected for execution.

According to an article published by New Scientist, computer scientist Stamos Karamouzis, and criminologist Dee Wood Harper at Loyola University in New Orleans, believe to have found scientific evidence that links inmate profiles to executions. While the two researchers said they did not find any specific patterns in a trial run of 3228 inmates, 53 of which were executed in 2006, they said that an artificial neural network (ANN) was able to come up with some keys.

In a second run, they used a computer system that removed obvious considerations. They trained the system with profiles of 1000 inmates between 1973 and 2000, consisting of 18 factors, including the inmate’s sex, age, race, marital status, educational level and information on their capital offences. Then they completed the database with 366 inmates and were wondering if the system would be able to predict executions. The result: The system was right in more than 90-percent of cases.

It turns out that certain profile data can give a clear indication whether a death row inmate will be executed or not. Gender was the most significant factor, as women are rarely executed. A clear indication was also the education level of the inmate, suggesting that the ability of an inmate to direct his appeal process can decide between life and death. The two scientists said that race was not found to be a decisive factor in execution decisions.

While the researchers do not expect their work to have much effect on policy, they believe that the predictions may have "serious implications" on the fairness of the justice system.