2010
DOI: 10.1177/1094428110366036
|View full text |Cite
|
Sign up to set email alerts
|

Method Variance and Marker Variables: A Review and Comprehensive CFA Marker Technique

Abstract: Lindell and Whitney introduced a partial correlation technique, now referred to as the correlational marker technique, for controlling method variance using a marker variable that is theoretically unrelated to substantive variables in a study. This article (a) first reviews their specific analysis plan, and then (b) reviews empirical studies that have followed all or part of this plan. The authors also (c) describe a structural equation method that has been applied to the analysis of marker variables and (d) r… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

23
922
8
8

Year Published

2012
2012
2023
2023

Publication Types

Select...
8
1

Relationship

0
9

Authors

Journals

citations
Cited by 1,110 publications
(1,035 citation statements)
references
References 65 publications
(136 reference statements)
23
922
8
8
Order By: Relevance
“…The composite reliabilities of the scales at both time one and time two calculated according to Fornell and Larcker (1981) are presented in the diagonal of Table 2; these are all above .70 and represent adequate internal reliability. We calculated the reliabilities due to common method using the unconstrained loadings on the latent marker factor for both time one and time two data-sets (see Williams, et al, 2010); for time one the reliability of the method loadings was only 0.02, and for time two it was 0.03. There is support, therefore, for the premise that common method variance was not inflating the reliability estimates, nor correlations between factors, nor factor loadings at either time one or time two.…”
Section: Resultsmentioning
confidence: 99%
“…The composite reliabilities of the scales at both time one and time two calculated according to Fornell and Larcker (1981) are presented in the diagonal of Table 2; these are all above .70 and represent adequate internal reliability. We calculated the reliabilities due to common method using the unconstrained loadings on the latent marker factor for both time one and time two data-sets (see Williams, et al, 2010); for time one the reliability of the method loadings was only 0.02, and for time two it was 0.03. There is support, therefore, for the premise that common method variance was not inflating the reliability estimates, nor correlations between factors, nor factor loadings at either time one or time two.…”
Section: Resultsmentioning
confidence: 99%
“…Without measurement invariance, (knowingly or unknowingly) researchers invoke assumptions about measurement equivalence in conducting tests of substantive hypotheses. Finally, Taylor and Ishida (2010) recommend the use of marker variables (Williams, Hartman, & Cavazotte, 2010) to assess potential common method variance associated with self-reports as a measurement model. Such variance can arise from respondent's consistency motifs, transient mood states, illusory correlations, item similarity, and social desirability.…”
Section: A Word About Measurement Methodsmentioning
confidence: 99%
“…The respondents group was regarded as credible since more than half of them have had at least 3-year of managerial experience. To further mitigate the potential common method bias (CMB), CFA marker technique (Williams, et al, 2010) has been performed on all 15 indicators. The test results indicate no significant presence of single common factor.…”
Section: Sampling and Data Collectionmentioning
confidence: 99%