A statistical problem that was impossible to solve just became possible

A common statistical method for estimating choices, called the maximum score estimator, used to be very hard to use because its results were unpredictable. New research shows that if the data has multiple layers of connections, this method now produces clear, predictable results. This means researchers can now use this method reliably to understand how people make choices, even with complex data.

For decades, economists and social scientists have struggled to use a powerful statistical tool because its outputs were too messy to interpret. This paper shows how to get clear answers from it, opening up new ways to analyze data where choices are involved, like in market research or policy evaluation. It means that a whole class of problems that were previously too complex for this method can now be tackled with more confidence.