Not all nonnormal distributions are created equal: Improved theoretical and measurement precision

Area of Science:

• Organizational Psychology
• Quantitative Psychology
• Behavioral Economics

Background:

• Understanding individual output distributions is crucial for performance analysis.
• Existing research often assumes specific distribution shapes (e.g., normal) or mechanisms.
• A comprehensive framework for classifying output distributions and their underlying causes is needed.

Purpose of the Study:

• To develop a four-category taxonomy of individual output distributions and their associated generative mechanisms.
• To introduce and implement 'distribution pitting,' a falsification-based method for comparing distribution fits.
• To identify the dominant distribution shapes and generative mechanisms in real-world occupational data.

Main Methods:

• Developed a taxonomy classifying distributions into pure power law, lognormal, exponential tail, and symmetric types.
• Associated each distribution type with a unique generative mechanism (self-organized criticality, proportionate differentiation, incremental differentiation, homogenization).
• Applied distribution pitting to 229 samples of individual output across diverse occupations.

Main Results:

• Exponential tail distributions, linked to incremental differentiation, were found to be the dominant shape in 75% of samples.
• This finding challenges prior conclusions about the prevalence of other distribution types and mechanisms.
• The study provides a new theoretical basis for understanding the link between past and future individual output.

Conclusions:

• Incremental differentiation is a prevalent mechanism shaping individual output across many occupations.
• The proposed taxonomy and distribution pitting methodology offer a robust framework for future research.
• Practical insights are provided for enhancing individual output and fostering top performance.