He developed tools for researchers to analyze categorical data, revolutionizing the study of poverty, income inequality and social mobility. He died of Covid-19.
Until Leo A. Goodman began his work on statistics in the early 1950s, researchers in the social sciences had a problem. It was easy enough to quantify the relationship between two numerical measurements — say, how height correlates to income level. But what about nonnumerical categories, like race and occupation?
There were statistical methods available from the natural sciences, like physics, but they were crude and imprecise when applied to population data. At the same time, postwar America was seeing a boom in data of all sorts: Census research, public polls, marketing surveys and mountains of information gleaned from the millions of men who had served in World War II.
It was a gold mine for sociologists, and Professor Goodman gave them the tools to dig into it.
He arrived at the University of Chicago in 1950 as a 22-year-old assistant professor of sociology and statistics, and almost immediately began churning out landmark papers that revolutionized both his fields. Over the course of his nearly 70-year career — he didn’t retire until 2017, when he was 89 — he developed not only the framework for analyzing huge sets of categorical data, but the statistical instruments to show relationships among those categories.
His work had an immediate and lasting impact on the study of subjects like poverty and social mobility. And as sophisticated quantitative analysis migrated into other fields, so did his methods: Today his influence can be felt in areas as distinct as management studies and computer science, where some of his statistical modeling tools are being applied to machine learning.
Professor Goodman died on Dec. 22 in a hospital in Berkeley, Calif. He was 92. The cause was complications of Covid-19, his son Andy said.
“Leo transformed the way categorical data is analyzed,” said Yu Xie, a sociologist at Princeton. “He was a genius, a legend.”
Leo Aria Goodman was born in Brooklyn on Aug. 7, 1928. His parents, Abraham Goodman and Mollie (Sacks) Goodman, were Ukrainian Jews who had immigrated to the Borough Park neighborhood, where Leo grew up. His father worked for his mother’s father, who owned a textile factory on the Lower East Side of Manhattan.
He was only 16 when he graduated from Stuyvesant High School in Manhattan; four years later he was valedictorian of his class at Syracuse University, where he majored in math and sociology. It took him just two years to complete his doctorate in math at Princeton.
“I can’t tell you how much he impressed us,” she wrote in a letter to Ms. Goodman in 1960. “So brilliant, kind, versatile and so very handsome. A match, a match.”
He and his wife later divorced. In addition to his son, he is survived by another son, Tom; his sister, Janice Towers; and five grandchildren.
Professor Goodman did much of his early work with William Kruskal, a mathematician at the University of Chicago. Three of the analytical tools they developed, and which carry their names — Goodman-Kruskal lambda, gamma and tau — are still widely used in statistical software.
He moved to the University of California, Berkeley in 1986, a few years after developing a rare form of cancer. His doctors wanted to amputate his right leg, but he dug into the medical literature and discovered that chemotherapy and new forms of therapy could save it.
Though the doctors ultimately removed the quadriceps in his leg, he was able to walk with barely a limp after several years of physical therapy. When colleagues would pass him on campus and ask how he was, he would reply, “Not good” — adding, after a beat, “Terrific!”
Source: New York Times