Statistics research: the next ten years

✍ Scribed by David S. Moore

• Publisher: Springer US
• Year: 1993
• Tongue: English
• Weight: 214 KB
• Volume: 3
• Category: Article
• ISSN: 0960-3174
• DOI: 10.1007/bf00141782

✦ Synopsis

I once spent an undergraduate summer with my left hand chained to a desk calculator, calculating double star orbits for a distinguished astronomer. Each lengthy calculation produced a point, which I plotted on graph paper. Theory required that these points, which contained a good deal of observational error, form a parabola. After some weeks of calculation, I therefore sharpened a pencil and drew a parabola through the scattered points by hand. The distinguished astronomer was pleased; he said it was clear that I was a mathematics student, because the parabola I drew 'looked like a parabola'. When I arrived at Purdue in 1967, I repeated this story to an older colleague. He replied that he had once visited some engineering faculty who also had a theory that demanded a parabola. They plotted their data on a large piece of paper, put it on the wall, and fitted a parabola by hanging a chain over the paper.