Comment on 'Anomalous Conductance Distribution in Q1D Gold Wires: Possible Violation of the One-Parameter Scaling Hypothesis' [PRL 88, 146601 (2002)]

TL;DR

This paper critiques a previous study claiming violation of the one-parameter scaling hypothesis in gold wires, demonstrating that flaws in data analysis invalidate those conclusions and reaffirming the hypothesis's validity.

Contribution

The authors identify errors in the previous analysis and provide corrected cumulant values, showing the original claims of violation are unsupported.

Findings

Corrected cumulant values are much smaller than previously claimed

Higher order cumulants are statistically unreliable due to limited data

The original data do not support violation of the one-parameter scaling hypothesis

Abstract

Mohanty and Webb [Phys. Rev. Lett. 88, 146601 (2002), cond-mat/0204298] claim that their data on conductance fluctuations in gold wires contradict the one-parameter scaling. We show that flaws in extracting values of the cumulants (irreducible moments) <<g^n>> of the conductance distribution (for n=3,4) invalidate all the conclusions made there. The actual values of the cumulants determined by us from the published raw data contained in Ref. [Phys. Rev. Lett. 88, 146601 (2002)] are orders of magnitude smaller than those claimed. Moreover, the limited applicability of the ergodicity hypothesis makes the higher order cumulants extracted from magnetofingerprints statistically unreliable. Thus, the data of Ref. [Phys. Rev. Lett. 88, 146601 (2002)] do not warrant any statement on the violation or validity of the one-parameter scaling.