The Weather--And More
No one can predict the future, of course, but statistical analysis can make estimates with varying levels of accuracy about the future. And we rely on these estimates all the time in our lives: the weather, risks associated with certain behaviors (smoking, eating fatty foods), the likelihood of a sporting team or political candidate winning, the ups and downs of the stock market. Still, as we go about our days it can be easy to forget these probabilities; our most profound interaction with statistics, after all, is usually deciding whether or not to bring an umbrella to work.
In business, however--and especially in the social impact / financial services business--we are constantly making guesses about the future. As a recent incident has highlighted, hiring an employee can be a crapshoot; when all is said and done, the interview process is all about reducing the risk of a bad hire to the lowest level possible. The challenge? People are complicated. They are hard to judge, and each person has different judgement. One person may seem lackluster on paper and phenomenal in-person, only to turn out to be unreliable and irresponsible. Another may receive a tepid letter of recommendation yet thrive in a particular role: maybe it's the new environment, maybe it's the tasks associated with the position, or maybe it's something else.
In the social sector in general, our work centers on guesses as to what product or service will have a positive impact on someone's life. Take our Financial Coaching: until the results of our Randomized Control Trial are in, we can't know for sure that, on the whole, it has the impact we are confident it has. What's more, every client is different--different life circumstances, job prospects, family situation, income, etc. Surely the same service can't work in the same way for everyone? Of course not, so we do our best to train our Financial Coaches to tailor the Coaching to the client's needs--again, making an educated guess as to what will make a difference.
In the financial services industry in particular, predicting the future is at the very core of what we do...A loan disbursal is little more than a prediction that someone will repay. We use a combination of quantitative and qualitative data to assess a client's ability to honor a loan, as well the likelihood that the debt will be honored. We then charge an interest rate designed to offset the margin of error in our calculation.
Who Pays Back?
But what's fascinated me over the years is trying to figure out what really motivates a person to pay us back. Yes, our repayment rates are phenomenal (over 90%), but we are still frequently surprised by the results. One borrower has a minimal history of paying back debts, appears to be unable to afford a loan, and ends up making every payment on-time! Another borrower has pretty good credit and seems responsible, and then disappears after one payment.
In short, the more I think about it, the more I realize that predicting the future features prominently in our lives. With the information around us, we make the most educated guesses we can and act accordingly. Of all the things I'd like to better understand, statistics and computer science are at the top of the list. The ability to look at a data set, slice-and-dice it and end up with useful statistics, seems to me to be a kind of wizardry, like stripping away the darkness of a cloud to know, with 90% certainty, how much rain will fall from it. And when I talk about data, I don't just mean numbers like credit scores, polling results and body fat. When I mean is that everything is data: the way a person smiles, the smell of breakfast, the tone of one's voice.
If everything is data, the only way we can keep our heads above water is to know how to sift through it and tease out the things that matter. In our case, what matters is that we change lives in a cost-effective manner. When we get it right, well what could be more magical than that?
--
Can you think of more ways in which statistics plays a role in your life? Let us know in the comments!
No one can predict the future, of course, but statistical analysis can make estimates with varying levels of accuracy about the future. And we rely on these estimates all the time in our lives: the weather, risks associated with certain behaviors (smoking, eating fatty foods), the likelihood of a sporting team or political candidate winning, the ups and downs of the stock market. Still, as we go about our days it can be easy to forget these probabilities; our most profound interaction with statistics, after all, is usually deciding whether or not to bring an umbrella to work.
In business, however--and especially in the social impact / financial services business--we are constantly making guesses about the future. As a recent incident has highlighted, hiring an employee can be a crapshoot; when all is said and done, the interview process is all about reducing the risk of a bad hire to the lowest level possible. The challenge? People are complicated. They are hard to judge, and each person has different judgement. One person may seem lackluster on paper and phenomenal in-person, only to turn out to be unreliable and irresponsible. Another may receive a tepid letter of recommendation yet thrive in a particular role: maybe it's the new environment, maybe it's the tasks associated with the position, or maybe it's something else.
In the social sector in general, our work centers on guesses as to what product or service will have a positive impact on someone's life. Take our Financial Coaching: until the results of our Randomized Control Trial are in, we can't know for sure that, on the whole, it has the impact we are confident it has. What's more, every client is different--different life circumstances, job prospects, family situation, income, etc. Surely the same service can't work in the same way for everyone? Of course not, so we do our best to train our Financial Coaches to tailor the Coaching to the client's needs--again, making an educated guess as to what will make a difference.
In the financial services industry in particular, predicting the future is at the very core of what we do...A loan disbursal is little more than a prediction that someone will repay. We use a combination of quantitative and qualitative data to assess a client's ability to honor a loan, as well the likelihood that the debt will be honored. We then charge an interest rate designed to offset the margin of error in our calculation.
Who Pays Back?
But what's fascinated me over the years is trying to figure out what really motivates a person to pay us back. Yes, our repayment rates are phenomenal (over 90%), but we are still frequently surprised by the results. One borrower has a minimal history of paying back debts, appears to be unable to afford a loan, and ends up making every payment on-time! Another borrower has pretty good credit and seems responsible, and then disappears after one payment.
In short, the more I think about it, the more I realize that predicting the future features prominently in our lives. With the information around us, we make the most educated guesses we can and act accordingly. Of all the things I'd like to better understand, statistics and computer science are at the top of the list. The ability to look at a data set, slice-and-dice it and end up with useful statistics, seems to me to be a kind of wizardry, like stripping away the darkness of a cloud to know, with 90% certainty, how much rain will fall from it. And when I talk about data, I don't just mean numbers like credit scores, polling results and body fat. When I mean is that everything is data: the way a person smiles, the smell of breakfast, the tone of one's voice.
If everything is data, the only way we can keep our heads above water is to know how to sift through it and tease out the things that matter. In our case, what matters is that we change lives in a cost-effective manner. When we get it right, well what could be more magical than that?
--
Can you think of more ways in which statistics plays a role in your life? Let us know in the comments!
No comments:
Post a Comment