Paper Type

Complete

Abstract

Are there examples of statistical techniques that are extremely popular in their usage by virtue of their inability to accomplish what they purport to do? This paper focuses on the old “Zombie” known as Harman’s single-factor test (HSFT) used as a diagnostic test for common method variance (CMV). Our literature review shows HSFT continues to proliferate in the Information Systems (IS) field with not a single instance of detecting CMV. To add to the evidence against HSFT, we conduct 2 Monte Carlo simulations. Our results show that the HSFT estimates are unreliable due to being affected by structural and empirical characteristics of the theoretical model (i.e., item reliability, correlational strength, and number of constructs) rather than simply the presence of CMV. These findings call into question conclusions drawn in numerous published studies and urge that researchers discontinue use of HSFT.

Paper Number

1885

Comments

NEXTTRANS

Download

Share

COinS
 
Aug 15th, 12:00 AM

Zombie Diagnostic Tests that Refuse to Die: The Harman’s Single-Factor Test

Are there examples of statistical techniques that are extremely popular in their usage by virtue of their inability to accomplish what they purport to do? This paper focuses on the old “Zombie” known as Harman’s single-factor test (HSFT) used as a diagnostic test for common method variance (CMV). Our literature review shows HSFT continues to proliferate in the Information Systems (IS) field with not a single instance of detecting CMV. To add to the evidence against HSFT, we conduct 2 Monte Carlo simulations. Our results show that the HSFT estimates are unreliable due to being affected by structural and empirical characteristics of the theoretical model (i.e., item reliability, correlational strength, and number of constructs) rather than simply the presence of CMV. These findings call into question conclusions drawn in numerous published studies and urge that researchers discontinue use of HSFT.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.