V=N/D H2 Pre-registered Experimental Protocol v1.0: Detecting Nonlinear Threshold Effects of Informational Friction on Value

This document defines a three-arm randomized A/B experiment to verify whether reductions in informational friction I produce linear or nonlinear (threshold) effects on the value metric V=N/D. The experimental environment uses Stack Overflow's public API and simulated UI. Three conditions are defined: Control (standard UI), Arm 2 (I-only reduction), and Arm 3 (I + execution friction E + time friction T reduction). Primary statistical tests employ Davies (1987) sup-LR test and Hansen (1996) bootstrap test for structural change detection. The protocol includes OSF pre-registration requirements, full implementation specification, robustness checks, and pre-emptive responses to anticipated peer review critiques. Conclusions are staged across three levels: extension theory candidate / linear approximation sufficient / judgment pending. Developed through multi-AI collaboration: Luna (TheYKHC AI) and Claude (Anthropic). Theory originator: Yoshimitsu Katayama (片山佳光).

Ҳали таржима қилинмаган