Conceptual Model Building: Overview Felicia Hill-Briggs, PhD, ABPP Associate Professor Departments of Medicine and Health, Behavior, and Society, Welch.

Conceptual Model Building: Overview

Felicia Hill-Briggs, PhD, ABPP

Associate Professor Departments of Medicine and Health, Behavior, and Society,

Welch Center for Prevention, Epidemiology, and Clinical Research

• Conceptual Model (Earp & Ennett, 1991):

– A diagram of proposed causal linkages among a set of concepts believed to be related to a particular public health problem

Often,• Integration of theories from multiple disciplines• Presentation of multi-level causality (e.g. individual,

group, society, environment; personal, economic, infrastructural, political)

Definition of Terms

• Model

– Visual representation of the elements of a theory – A conceptual framework for organizing and integrating

information– A conceptual structure successfully developed in one

field and applied to some other field to guide research and practice

(Marx MH & Goodson FE. Theories in Contemporary Psychology, 1976)

1. To organize and synthesize related factors into a coherent, simplified representation

2. To make alternative routes to an endpoint explicit

3. To summarize an existing body of literature and/or propose new research directions

• Guides formulation of hypotheses for research questions• Helps with planning data collection (confounders,

moderators, mediators) and data analyses

Conceptual Model Purposes and Utility

Tips for Building Conceptual Models

• Remember that models are quite individualized; there is no one “right” representation.

• Consider that relationships depicted are driven by a combination of theory and evidence.

• Start with a basic, simple model structure and add additional sets of variables as needed

• Determine your desired level of variable specificity, based on purpose of your model, clarity regarding relationships among variables, measurement that may be employed.

– Generally, values for variables should not be included (e.g. age >65, disease duration <10 years).

• Share your model with colleagues both knowledgeable and unknowledgeable about your content area for feedback.

Considerations

Accurate Representation of the Science or Theory

• What are the relationships among the variables depicted?– Independent, dependent– Antecedant, predictor– Moderators, mediators

• Is there an ordering of the variables (e.g. rank, sequence or temporal)

• Are there interactions among sets of variables?

Stylistic Elements for Communicating Relationships and Emphasis

• Use of different shapes and object sizes; placement

• Use of border styles

• Use of lines (styles, sizes) to denote connectivity

• Use of arrows (styles, sizes) for directionality

• Use of emphasized text (bold, underline, case)

Examples of Basic Diagrams

IOM Committee on the Future of Emergency Care in the US Health System, 2004

Cooper. JGIM 2002

Biessels. Lancet Neurol 2006

Informed,ActivatedPatient

ProductiveInteractions

Prepared,ProactivePractice Team

DeliverySystemDesign

DecisionSupport

ClinicalInformation

Systems

Self-Management

Support

Health System

Resources and Policies

Community

Health Care Organization

Chronic Care Model

Improved Outcomes

Institute of Medicine

Choosing the Scope of the Model

• Comprehensive conceptual framework representing a field of study

• Study-specific conceptual model

– Focus on one variable or a subset of variables in the comprehensive framework and elaborate beyond the scope of the comprehensive framework

• Inclusion of specific variables to be measured

– Adapt comprehensive framework to address selected variables throughout the comprehensive model

Fisher. Diab Educ 2007