Statistical theories of success

Area of Science:

• Bibliometrics
• Scientometrics
• Sociology of Science

Background:

• The Pareto and log-normal distributions are frequently employed to model the statistical distribution of success.
• Lotka's observations on publication rates provide extensive data for analyzing scientific success.
• Existing models often fail to capture the full spectrum of success distribution.

Purpose of the Study:

• To compare the efficacy of Pareto and log-normal distributions against Lotka's publication data.
• To introduce and validate a new flat factor analysis model for performance.
• To determine the number of environmental factors influencing scientific success.

Main Methods:

• Comparative analysis of Pareto and log-normal distributions using Lotka's publication data.
• Application of a recently developed flat factor analysis technique.
• Examination of success distribution across low, moderate, and high levels of achievement.

Main Results:

• Pareto and log-normal distributions provide adequate fits only for low to moderate success rates.
• The flat factor analysis demonstrates excellent agreement with observed success distribution across all levels.
• Lotka's data, analyzed via flat factor analysis, allows for the estimation of environmental factors.

Conclusions:

• Traditional distributions are limited in their ability to model the complete range of scientific success.
• Flat factor analysis offers a superior and comprehensive approach to understanding performance and success.
• The study provides a framework for identifying and quantifying environmental influences on scientific output.