Thursday, September 26, 2002

This presentation is part of : Posters

Sustaining Scientific Integrity in Interactive Research Projects

Donna L Algase, RN, PhD, FAAN, FGSA, professor, Division of Acute, Critical and Long Term Care, University of Michigan, Ann Arbor, MI, USA, Ann Kolanowski, RN, PhD, professor, School of Nursing, Penn State University, State College, PA, USA, Ann Whall, RN, PhD, FAAN, professor, School of Nursing, Division of Acute, Critical & Long-term Care, University of Michigan, Ann Arbor, MI, USA, Cornelia Beck, RN, PhD, FAAN, professor, University of Arkansas for Medical Sciences, Little Rock, AR, USA, and NDB Research Group, University of Michigan and University of Arkansas for Medical Sciences, USA.

Objective: This poster articulates an approach to maintaining scientific integrity in interactive research projects.

Design: The Need-driven Dementia-compromised Behavior (NDB) model is used as a template in a correlational design driving three inter-related studies on wandering, aggression, and problematic vocalizations in nursing home residents. Each study is replicated in an independent sample drawn from the other projects. Data from all three projects are further combined to test the power of the NDB model to explain any dementia-related behavior.

Population: Subjects with a documented dementia, MMSE score below 24, and meeting other project-specific criteria designed to eliminate confounds are recruited from among approximately 30 nursing homes. Strict inclusion and exclusion criteria are used to screen subjects for eligibility within and across projects. Approximately 850 subjects are sought across all three projects.

Concepts: Because projects are interactive, threats to validity of research design arising in one project (and remedies that might be applied) can have adverse implications for other projects. In these studies, variation in samples and sampling designs necessary to meet the aims of individual projects presented considerable challenges to maintaining scientific integrity for interactive research goals. This poster delineates threats to construct and external validity encountered and describes strategies used to minimize such threats.

Findings: Carefully articulated decision trees for subject enrollment will demonstrate our efforts to control threats to external validity. We will also illustrate how an integrated event-specific and time-sampling design was used to overcome threats to construct validity.

Conclusions: Interactive research studies introduce a level of complexity that requires consideration of validity of design within, between, and across all projects. While challenging, interactive research studies are a powerful means to advance nursing science in a delineated area of study and, thus, warrant the effort.

Back to Posters
Back to The Advancing Nursing Practice Excellence: State of the Science