Earlier this week, Dr. John Shieh of the Memorial University Department of Computer Science came under fire for an assignment dealing with a particularly uncomfortable topic. In the assignment, students were introduced to a hypothetical woman named Heather, who was a rape victim. The computer program the students were meant to develop was a predictive model over whether Heather was likely to commit suicide.
On the surface, that’s never a comfortable question to ask. But, as a computer engineer, I would’ve gladly done the assignment.
Predictive statistics for victims of traumatic events are very crucial to their long-term care and treatment. They’re often used over the course of decades to determine if present measures of therapy and other aid are having the desired effect. Many studies show unsurprisingly that victims of rape, assault, and other personal crimes have an unfortunately high suicide rate. The public reaction just feels like one that is very much tied to the use of the word “rape.” If the question were instead framed as predicting the likelihood of a person who had previously attempted suicide to attempt to take their life a second time, would we be instead praising the relevance and importance of the question?
I have had to deal with the issue of sexual assault and rape in a very personal way with people very close to me. If I knew that they were in the class, I would be even more compelled to tackle the assignment in hopes that I might one day provide the necessary knowledge needed to get them help and to ensure they could live a life as normal as they had before said event.
Rape is a serious issue. This very fact underscores why we need to tackle it to the best of our ability. We need to have serious dialog, and the kind of solution development that can only be gained through team discussions and projects just like these. And it starts with being able to ask the question.
These are the kinds of tough questions with ill-defined parameters and solutions that we should be posing to future computer scientists. These are the kind of questions that make a difference in the world. Not the statistics of drawing a blackjack or the number of jelly beans likely to be in a jar. Though safe, these questions would not develop the kind of critical lens and abstracting process required for solving a real statistical problem backed with computerized informatics.
Shieh has come under fire for being insensitive and, amazingly, even been used as an example of why sexual harassment needs to be better understood by the public. The truth is, Shieh is courageous to offer a real-world uncomfortable issue to his students and ask: “Now what can you do to help?”