Multivariate Research Methodology

Henson, Robin K.; Hinerman, Krystal M.

doi:10.1093/obo/9780199756810-0145

Cited by 1 publication

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, item skewness for ATCB ranged from −1.53 to −0.93, and item kurtosis ranged from 0.68 to 3.32. However, univariate normality is a necessary but not sufficient condition for multivariate normality (Henson, 1999). By using Mplus 8.2 software (Muthén & Muthén, 2018) it was determined that the data were not multivariate normal as the scaling correction factor was 1.578.…”

Section: Methodsmentioning

confidence: 99%

“…The skewness and kurtosis of these 11 items ranged from −1.27 to −1.00 and from 1.57 to 3.25, respectively. However, univariate normality is only a necessary but not sufficient condition for multivariate normality (Henson, 1999) so Mardia's coefficient of multivariate kurtosis was calculated using the DeCarlo (1997) macro. The normalized (i.e., standardized) coefficient was 163.55.…”

Section: Subsample a Proceduresmentioning

confidence: 99%

See 1 more Smart Citation

Attitude Toward the Color Blue: An Ideal Marker Variable

Miller

Simmering

2022

Organizational Research Methods

View full text Add to dashboard Cite

Researchers often turn to post hoc statistical techniques to identify common method variance (CMV) in same source data and one viable option is to use a marker variable. The choice of marker variable is important, yet these variables are difficult to find, primarily because they must be theoretically unrelated to study variables but measured in the same way (e.g., perceptual; on a Likert scale). This manuscript uses scale development best practices to create a marker variable—attitude toward the color blue—that can be applied in a wide variety of social science research. Scale reliability and validity are addressed, discriminant validity with other measures that detect CMV is tested, and the Confirmatory Factor Analysis Marker Technique is applied with this scale. An experiment designed to analyze the effect of the placement of the scale in surveys is reported. Recommendations to researchers for use of this new scale to detect CMV are provided.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Subsample a Proceduresmentioning

confidence: 99%