Dispelling Four Myths about Correction for Attenuation in Communication Trait Research

Abstract

Four popular myths about correction for attenuation are dispelled because (a) the correction for attenuation protocol can increase the amount of error in a study, (b) correcting for attenuation should not be used to reduce the risk of Type II error (Type II errors are preferable to Type I errors), (c) scores corrected for attenuation cannot improve predictive validity, and (d) scores corrected for attenuation are not easily interpretable when considering typical research results. Correction for attenuation is valuable for stimulating research and bolstering sample sizes when used in meta-analyses. Most of the problems with the procedure come from difficulties in interpretation. Limitations on the procedure underscore the need to work toward improving measures of communication traits to make them more reliable, more valid, and, ultimately, more useful in explaining and predicting outcome variables.

Keywords:

• Correction For Attenuation
• Research Methods
• Traits

Additional information

Notes on contributors

Charles J. Wigley

Charles J. Wigley III (PhD, Kent State University, 1986) is a professor in the Department of Communication Studies at Canisius College.