You may not realize the extent to which predictive modeling, the process of forecasting future outcomes, is being used in the marketplace. According to IKO System, several familiar companies have come to depend on predictive analytics.
Netflix uses modeling to predict which movies and shows you most want to watch. Facebook predicts preference patterns based on “likes” and, according to their data scientists, can even predict the longevity of romantic relationships. Match.com goes even further with an algorithm that predicts possible matches.
Predictive modeling affects your health care. The Harvard Business Review reported in 2016 that the Parkland Health and Hospital System in Dallas, Texas uses an algorithm to predict a patient’s readmission risk with heart failure. The Veteran’s Health Administration, they go on to say, has collected three decade’s worth of electronic data from its patients. And Kaiser Permanente in California has used predictive analysis to reduce the overuse of antibiotics in newborns.
Predictive modeling may not always be a matter of life and death, but it’s certainly more ubiquitous now. So where did it all start? It actually began with two friends who just happened to be geniuses.
In 1973, Amos Tversky and Daniel Kahneman, two researchers at Hebrew University in Jerusalem, were working out of the Oregon Research Institute when they wrote a landmark paper, “On the Psychology of Prediction,” and published it in the Psychological Review. Certainly there had been work in prediction before, but never did the work point to an idea that disrupted the way people viewed themselves. We like to think of ourselves as rational — and so did economic theorists — weighing our choices and making rational decisions. As it turns out, Tversky and Kahneman proved this soundly to be untrue.
“In making predictions and judgments under uncertainty,” said Tversky and Kahneman, “people do not appear to follow the calculus of chance or the statistical theory of prediction. Instead, they rely on a limited number of heuristics which sometimes yield reasonable judgments and sometimes lead to severe and systematic errors.”
In short, people sometimes make decisions irrationally.
At this point, you may be asking, How did they arrive at these findings? In fact, who are Amos Tversky and Daniel Kahneman? And what in the world is a ‘heuristic?’ All of this is answered in Michael Lewis’s new book, The Undoing Project: A Friendship That Changed Our Minds. The book is as much a biography of a friendship between Tversky and Kahneman as the record of their development of the idea of heuristics, which are systematic biases inherent to the human mind that often make us act against our own best interests. After Tversky died, Kahneman went on to win a Nobel Prize in Economics.
Why economics? Because their findings suitably applied to the choices people routinely make in purchasing, banking, investing, and so on. It disabled the view that people make informed choices and led to the greater use of statistics — rather than intuition — in making important decisions. Their work essentially created the field of behavioral economics, which incorporates psychology into economic decision-making — not only for individuals but for entire institutions.
Naturally, their research lends itself to the Big Data revolution. To power these predictive models, the model needs data. Every time you make a choice on what movie to watch, you teach Netflix a bit more. When Facebook sees that you’re single, you start getting ads for dating apps. When you find a profile on Match.com that you prefer, you will begin to see more like it.
Capture, too, falls into this category. Prospective students also create data; they have preferences and profiles. Using this data, we create a predictive model via Envision in order to find them the right “match” as well — the college that most fits their needs. There’s reason to be confident in the work of Tversky and Kahneman. A Nobel Prize suggests, at the very least, that predictive modeling is on to something.
By Sean Hill, Senior Content Writer, Capture Higher Ed