<p>The existence of <i>jingle fallacies</i> (equally named constructs/measures that, in fact, assess different constructs) and <i>jangle fallacies</i> (differently named constructs/measures that, in fact, measure the same concept) jeopardizes psychological assessment, as both are associated with conceptual and assessment-related uncertainties. A guideline presented by Lawson and Robins <i>Personality and Social Psychology Review</i>, <i>25</i>, 344–366, (<CitationRef CitationID="CR12">2021</CitationRef>) helps evaluate the intensity of respective fallacies. While the guideline is well elaborated, psychometric aspects regarding (dis)similarities of nomological networks require extensions and differentiations. I recommend two analytical advancements, namely (a) the derivation of correlation difference hypotheses for criteria with which the allegedly jingled (jangled) variables are assumed to be correlated at equal (different) levels and (b) procedures to derive cutoffs for the overall similarity of nomological networks based on the <i>elemental approach</i> (Kay &amp; Arrow <i>Social and Personality Psychology Compass</i>, <i>16</i>, e12662, <CitationRef CitationID="CR10">2022</CitationRef>). Considering correlation difference tests, I further outline the importance of power analyses. These extensions help improve the evaluation of assumed jingle and jangle fallacies, arguably increasing the stability and reliability of research findings.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Extending Lawson and Robins’ (2021) guideline for the evaluation of jingle and jangle fallacies

  • Christian Blötner

摘要

The existence of jingle fallacies (equally named constructs/measures that, in fact, assess different constructs) and jangle fallacies (differently named constructs/measures that, in fact, measure the same concept) jeopardizes psychological assessment, as both are associated with conceptual and assessment-related uncertainties. A guideline presented by Lawson and Robins Personality and Social Psychology Review, 25, 344–366, (2021) helps evaluate the intensity of respective fallacies. While the guideline is well elaborated, psychometric aspects regarding (dis)similarities of nomological networks require extensions and differentiations. I recommend two analytical advancements, namely (a) the derivation of correlation difference hypotheses for criteria with which the allegedly jingled (jangled) variables are assumed to be correlated at equal (different) levels and (b) procedures to derive cutoffs for the overall similarity of nomological networks based on the elemental approach (Kay & Arrow Social and Personality Psychology Compass, 16, e12662, 2022). Considering correlation difference tests, I further outline the importance of power analyses. These extensions help improve the evaluation of assumed jingle and jangle fallacies, arguably increasing the stability and reliability of research findings.