Improving Precision and Confidence of Research Application Through Mediator and Moderator Models

Abstract

Purpose:

The purpose of this article is to conceptually discuss moderator and mediator models in healthcare design studies.

Background:

Healthcare outcomes are many times the result of meaningful interactions between multiple factors from different domains such as physical design, operations, culture, psychology, physiology, and social, among others. Simple multivariate models may not (a) capture these intricate relationships, (b) provide predictive precision, and (c) optimize arguments behind design decisions.

Methods:

A review of text books and publications on research methods was conducted, along with a review of completed research studies to identify appropriate examples for articulation.

Results:

An analysis of a nonrandom sample of completed healthcare design studies demonstrates the potential benefits of developing and testing moderator and mediator models in healthcare design studies.

Conclusions:

Moderator and mediator models may help optimize multifaceted solutions that include simultaneous and parallel interventions in the physical, operational, cultural, psychological, physiological, social, and/or other environments hypothesized to play a role in the outcome of interest.