Linear units improve articulation between social and physical constructs: An example from caregiver parameterization for children supported by complex medical technologies

2016 Joint IMEKO TC1-TC7-TC13 Symposium: Metrology Across the Sciences: Wishful Thinking?

Bezruczko, N.^1 ; Stanley, T.^2 ; Battle, M.^3 ; Latty, C.^3

Chicago School of Professional Psychology, Chicago, United States^1

Indiana University Health, Indianapolis, United States^2

Riley Hospital at Indiana University Health, Indianapolis, United States^3

关键词: General linear modeling;    Hierarchical analysis;    International researches;    Medical technologies;    Research programs;    Science methods;    Scientific paradigm;    Scientific researches;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/772/1/012020/pdf
DOI  :  10.1088/1742-6596/772/1/012020

来源: IOP

【 摘 要 】

Despite broad sweeping pronouncements by international research organizations that social sciences are being integrated into global research programs, little attention has been directed toward obstacles blocking productive collaborations. In particular, social sciences routinely implement nonlinear, ordinal measures, which fundamentally inhibit integration with overarching scientific paradigms. The widely promoted general linear model in contemporary social science methods is largely based on untransformed scores and ratings, which are neither objective nor linear. This issue has historically separated physical and social sciences, which this report now asserts is unnecessary. In this research, nonlinear, subjective caregiver ratings of confidence to care for children supported by complex, medical technologies were transformed to an objective scale defined by logits (N=70). Transparent linear units from this transformation provided foundational insights into measurement properties of a social- humanistic caregiving construct, which clarified physical and social caregiver implications. Parameterized items and ratings were also subjected to multivariate hierarchical analysis, then decomposed to demonstrate theoretical coherence (R2>.50), which provided further support for convergence of mathematical parameterization, physical expectations, and a social-humanistic construct. These results present substantial support for improving integration of social sciences with contemporary scientific research programs by emphasizing construction of common variables with objective, linear units.