Mendel's Data

blueline

Topics Addressed

  1. Estimation of Proportions
  2. Genetics
blueline

The person who founded modern genetics was Gregor Mendel (for more information on Mendel and his work you can visit MendelWeb), who in a series of experiments on garden peas in about 1865 provided a scientific explanation for heredity and eventually caused a revolution in human understanding of life.

However, on the basis of a reanalysis of Mendel's data, a statistician named Fisher some 70 years later found compelling evidence that someone had fudged Mendel's data so that the agreement between his genetic theory and the outcome of his experiments was much too close for comfort. Mendel looked at a number of characteristics of pea plants, including their seed color and whether the plants flowered early or late in the season, and one of the things he noticed was that pea seeds are always either smooth or wrinkled. He bred a pure smooth strain and a pure wrinkled strain, resulting in first-generation hybrids, which all turned out to be smooth. He then crossed the first- generation hybrids with themselves to get second-generation hybrids, some of which were smooth and some wrinkled. Across his many field trials he put together data on 7324 second-generation hybrid plants, of which 5474 turned out to be smooth and 1850 wrinkled. Make up a simple genetic model to account for these results.

Under this model, what is the chance of agreement between the observed and expected frequencies of smooth and wrinkled as close as that reported by Mendel, or closer?
What do you conclude on the basis of this one set of results about whether he fudged his data?

Explain briefly.

blueline

George Michailides
gmichail@stat.ucla.edu