Top

Site Menu
History of Statistics

Early Data Analysts

Adolphe Quetelet (1796-1874)

Adolphe Quételet by Joseph-Arnold Demannez
Adolphe Quetelet (1796-1874)

Adolphe Quetelet was born in Ghent in modern-day Belgium. After his father died in 1803, Quetelet's family faced financial difficulties. So when he finished secondary school, he immediately began teaching. He completed a dissertation to earn a doctoral degree in 1819 and was appointed professor in Brussels. He felt an urge to help develop Brussels, inspiring him to seek and obtain approval for founding a city observatory, and travel to Paris to learn how to set it up. He shadowed the director of the Paris observatory, Alexis Bouwing, but also met Laplace, Poisson, and Fourier who introduced him to probability theory and its importance.

Quetelet returned to Brussels and began constructing the observatory, which became part of a bigger plan. He wanted to help expose the public to science and educate them in physics and astronomy. He organized public lectures at the Brussels Museum. In addition to his contributions in probability and statistics, he had many interests including art, poetry, literature, and languages.

One of Quetelet's major contributions to statistics is using probability theory to investigate human populations. Beyond what previous researchers had done, he examined not only the average, but how data varied from average. He found normal distributions in the data he observed such as heights and chest measurements. However, looking at data for military draft applications, he recognized fraud because there were too few height measurements just above 62 inches and too many just below. He noted that it was likely because the cutoff for mandatory service was 62 inches. This meant that many who measured just above 62 inches tall had bent their knees or in some other way adjusted so they would not be required to serve in the military. Quetelet recognized that we should expect randomness and use missing randomness to point out where something might be questionable. This is an idea which is now used in a field called forensic economics.