2009
DOI: 10.1177/1094428109351241
|View full text |Cite
|
Sign up to set email alerts
|

Common Method Bias in Regression Models With Linear, Quadratic, and Interaction Effects

Abstract: This research analyzes the effects of common method variance (CMV) on parameter estimates in bivariate linear, multivariate linear, quadratic, and interaction regression models. The authors demonstrate that CMV can either inflate or deflate bivariate linear relationships, depending on the degree of symmetry with which CMV affects the observed measures. With respect to multivariate linear relationships, they show that common method bias generally decreases when additional independent variables suffering from CM… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

65
1,422
8
11

Year Published

2012
2012
2023
2023

Publication Types

Select...
10

Relationship

0
10

Authors

Journals

citations
Cited by 2,238 publications
(1,560 citation statements)
references
References 24 publications
65
1,422
8
11
Order By: Relevance
“…Haas & Hansen, 2007). Siemsen et al (2009) also report that quadratic effects are very unlikely to be the result of common method bias and such effects -on the contrary -are more difficult to detect under circumstances of method bias. They conclude that "research articles whose primary purpose it is to examine quadratic effects should not be criticized for suffering from CMV" (Siemsen et al, 2009: 13).…”
Section: Control Variablesmentioning
confidence: 99%
“…Haas & Hansen, 2007). Siemsen et al (2009) also report that quadratic effects are very unlikely to be the result of common method bias and such effects -on the contrary -are more difficult to detect under circumstances of method bias. They conclude that "research articles whose primary purpose it is to examine quadratic effects should not be criticized for suffering from CMV" (Siemsen et al, 2009: 13).…”
Section: Control Variablesmentioning
confidence: 99%
“…However, there are several reasons that common method bias is likely not an issue in this research: the VIF values were not concerning, there were no high correlations, there was a time lag in measuring individual innovation, and the constructs were different from each other. Moreover, it has been stressed that if interaction relations are found in data, common method bias is unlikely an issue (Siemsen et al, 2010). Nevertheless, future research could use multisource data.…”
Section: Limitations and Future Researchmentioning
confidence: 99%
“…Once again, all the correlations between the dependent and the independent variables retained their significance after controlling for method bias. Finally, we applied the regression-based marker variable technique (Siemsen, Roth, & Oliveira, 2010). Accordingly, we included the marker variable (the percentage of time employees spend dealing with patients on a daily basis) in the regressions estimated to test the research hypotheses and observed that, after controlling for method bias, all substantive relationships remained statistically significant.…”
Section: -----------------------------Insert Table 2 About Here -----mentioning
confidence: 99%