Affordable Access

Structural equation model testing and the quality of natural killer cell activity measurements

Authors
Publisher
BioMed Central
Publication Date
Source
PMC
Keywords
  • Research Article

Abstract

1471-2288-5-1.fm ral BMC Medical Research ss BioMed CentMethodology Open AcceResearch article Structural equation model testing and the quality of natural killer cell activity measurements Leslie A Hayduk*1, Hannah Pazderka-Robinson2, Greta G Cummings3, Merry-Jo D Levers3 and Melanie A Beres1 Address: 1Department of Sociology, University of Alberta, Edmonton, Alberta, T6G 2H4 Canada, 2University Centre for Neuroscience, University of Alberta, Edmonton, Alberta, T6G 2S2 Canada and 3Faculty of Nursing, University of Alberta, Edmonton, Alberta, T6G 2G3 Canada Email: Leslie A Hayduk* - [email protected]; Hannah Pazderka-Robinson - [email protected]; Greta G Cummings - [email protected]; Merry-Jo D Levers - [email protected]; Melanie A Beres - [email protected] * Corresponding author Abstract Background: Browne et al. [Browne, MacCallum, Kim, Andersen, Glaser: When fit indices and residuals are incompatible. Psychol Methods 2002] employed a structural equation model of measurements of target cell lysing by natural killer cells as an example purportedly demonstrating that small but statistically significant ill model fit can be dismissed as "negligible from a practical point of view". Methods: Reanalysis of the natural killer cell data reveals that the supposedly negligible ill fit obscured important, systematic, and substantial causal misspecifications. Results: A clean-fitting structural equation model indicates that measurements employing higher natural-killer-cell to target-cell ratios are more strongly influenced by a progressively intrusive factor, whether or not the natural killer cell activity is activated by recombinant interferon γ (rIFN γ). The progressive influence may reflect independent rate limiting steps in cell recognition and attachment, spatial competition for cell attachment points, or the simultaneous lysings of single target cells by multiple natural killer cells. Conclusions: If the progressively influential factor is ultimately id

There are no comments yet on this publication. Be the first to share your thoughts.