Friday, December 26, 2008

Statistical analysis of Bible verse lengths


In the course of my Go Bible activities, I found it necessary to perform a statistical analysis of Bible verse lengths, in order to test a conjecture that there had been systematic truncation in the particular source text that I had been working with. This was the French translation by Pirot and Clamer.

This is an illustration of where my Engineering and Theology interests and skills overlap.

The underlying probability distribution would be of interest to statisticians.
  • Cubic polynomial best fit below the mode, where mode length = 79 characters
  • Linear best fit above the mode
The chart lends weight to my truncation conjecture. There was a second peak of verses with length=255. The few slightly longer verses may be explained by word-wrap effects.

Further details are given in this Go Bible Forum topic. If you are not a forum member, you will first need to register before logging in.