R E P O RT

A Logic Model Framework for Community Nutrition Education

LY D I A C. M E D E I RO S , P H D, RD 1 ; S U E N I C H O L S O N B U T K U S , P H D, RD 2 ;
H E L E N C H I P M A N , P H D, RD, LN 3 ; RU B Y H. C OX , P H D, RD 4 ; L A R RY J O N E S , P H D 5 ;
DEBORAH LITTLE, EDD6
1
Department of Human Nutrition,The Ohio State University, Columbus, Ohio;
2
Department of Food Science and Human Nutrition,Washington State University, Puyallup,Washington;
3
Cooperative Extension Service, South Dakota State University, Brookings, South Dakota;
4
Department of Human Nutrition, Foods and Exercise,Virginia Polytechnic Institute and State University,
Blacksburg,Virginia; 5Cooperative Extension, University of Wisconsin-Extension, Madison,Wisconsin;
6Mississippi State University Extension Service, Mississippi State, Mississippi

comes. For example, the University of Wisconsin-Extension

ABSTRACT and Ohio State University Extension programs use a logic
model for their annual reporting systems to consolidate out-
Logic models are a practical method for systematically col- comes from all statewide programs.The US Centers for Dis-
lecting impact data for community nutrition efforts, such as
ease Control and Prevention has developed a logic model for
the Food Stamp Nutrition Education program. This report
describes the process used to develop and test the Commu- education programs that teach physical activity to show
nity Nutrition Education Logic Model and the results of a comparable impact from program to program.1 Similar pro-
pilot study to determine whether national evaluation data gram planning and evaluation models are used by nonprofit
could be captured without losing flexibility of programming organizations, such as the United Way of America.2
and evaluation at the state level. The objectives were to Food Stamp Nutrition Education (FSNE), funded in part
develop an evaluation framework based on the Logic Model by the Food and Nutrition Service (FNS), US Department of
to include dietary quality, food safety, food security, and Agriculture (USDA), is a nationwide nutrition education
shopping behavior/food resource management and to effort that is administered and implemented by different state
develop a training mechanism for use.The portability feature agencies. FSNE operates under approved state plans to target
of the model should allow application to a variety of com- Food Stamp Program participants and applicants with basic
munity education programs.
food and nutrition education.3 When programs are imple-
KEY WORDS: logic model, nutrition education, food mented in widely different situations (eg, different states
stamps, diet quality, food safety, food security, food resource implementing the same or different curricula, different agen-
management cies developing programs within the same topic area), national
impact is more easily identified and credibility is enhanced
(J Nutr Educ Behav. 2005;37:197-202.) when a common outcome and indicator system is used.
A system for measuring outcomes across all FSNE pro-
jects, based on common goals and objectives, allows stake-
holders to determine if programs and social marketing cam-
INTRODUCTION paigns are performing as expected. The need for such a
system grew from professional discussions that resulted in a
Logic models have been adopted as program planning and comprehensive review of evaluation methodology used in
evaluation tools by some state Extension systems and other nutrition education.4 Subsequently, the FNS and the Eco-
community service providers as a practical method for sys- nomic Research Service (ERS) of the USDA sponsored the
tematically collecting impact data and summarizing out- development of “white papers” to review evaluation
methodology commonly used in 5 areas: systems and envi-
Funding was provided by Extension and Land-Grant Universities across the nation ronmental change, food shopping practices, food safety,
and from the USDA CSREES, Families, 4-H and Nutrition Unit. dietary quality, and food insecurity and security.The papers
Address for correspondence: Lydia C. Medeiros, PhD, RD, Department of Human
Nutrition,The Ohio State University, 1787 Neil Ave, 331A Campbell Hall, Colum- were published as a special issue of the Journal of Nutrition
bus, OH 43210-1295; Tel: (614) 292-2699; Fax: (614) 292-8880; E-mail: Education.5-9 Major concepts addressed in these papers were
[email protected].
©2005 SOCIETY FOR NUTRITION EDUCATION
linked in a model for nutrition education that illustrates rela-
197
198 Medeiros et al/COMMUNITY NUTRITION EDUCATION LOGIC MODEL

tionships of model components to the health and indepen-

dence of participants (Figure 1).
This baseline work was the inspiration for the development
of the Community Nutrition Education Logic Model (CNE
Logic Model).The goal of the CNE Logic Model is to guide
the development and evaluation of nutrition education pro-
grams implemented through a variety of sponsoring agencies.
The potential for the model’s use across community systems
beyond FSNE became apparent during the development Figure 2. Relationships of logic model components.
process, so early steps were taken to ensure the portability of
the model across programs.This report describes the process
and philosophy used to develop and test the CNE Logic facilitate consistent reporting across programs and projects. For
Model and outcomes from the first year of implementation. the CNE Logic Model, short-term outcomes and indicators
were defined as those that reflect increased awareness, knowl-
edge, and skills. Medium-term outcomes and indicators were
DESCRIPTION OF THE deemed those that reflected action taken and changed behav-
LOGIC MODEL CONCEPT iors. Long-term outcomes and indicators were tied to personal
health and nutritional independence factors that could be mea-
There are 3 major components of logic models (Figure 2): sured by research, population data, and public policy outcomes.
inputs, outputs, and outcomes that are developed in the con-
text of a program’s assumptions and external factors.10 In all
logic models, inputs are the resources that are made available METHOD USED TO DEVELOP
for the planning, implementation, and evaluation of projects; THE CNE LOGIC MODEL
examples of inputs are human resources, funding, facilities,
equipment, and curricula. Outputs are divided into activity A committee was convened in 2001 to address issues of pro-
and participant categories. Activity outputs are the educa- gram evaluation for FSNE. The committee was composed
tional program component (eg, classes, newsletters, demon- of 9 Cooperative Extension Service nutrition education
strations), and participant outputs are the demographics of specialists and leaders. One individual served as the com-
the individuals, families, and communities who are reached. mittee chair, and 4 subgroups of 2 members each were
Outcomes describe the changes or impacts that occurred formed to address the various white papers.5-9 The com-
because of the program outputs. Impacts can be evident as an mittee was charged with 2 objectives: to develop an evalu-
immediate result of the program (short term) or may not be ation system for FSNE based on the 5 evaluation white
apparent for some period of time in the future (medium papers with a logic model format and to develop a Web-
term).The impact on society as a whole may not be revealed based interface for training and gathering evaluation data
in the evaluation of individual programs but may be evident from FSNE across the nation.The reason for developing the
from nationwide data. The possibility of showing societal system was to track comprehensive and longitudinal nutri-
impacts (long term), collectively documented across a coor- tion education accomplishments and to facilitate commu-
dinated system of programs, further justifies the need for a nication of outcomes to a variety of local, state, and national
national system of program evaluation.11 stakeholders. The implementation and training of profes-
Change most likely occurs on a continuum and not in dis- sionals in the use of the model were the final steps of the
crete intervals.Yet recognizable cutoff points are necessary to process.The logic model format was selected as the frame-
work for the evaluation system.
The committee agreed that attainment of personal health
and the ability to independently provide for nutritional needs
are the ultimate goals of nutrition education (see Figure 1).
This is an adaptation of a previous model of nutrition educa-
tion.12 Many antecedent processes are necessary for an indi-
vidual to accomplish behavior change and achieve these 2
goals. However, for the purposes of reporting the impact of
FSNE, demonstrated skills and changed behavior are the mea-
surable outcomes because of their clear relationship to the
attainment of personal health and nutritional independence.
The CNE Logic Model (Figure 3) was based on 3
assumptions. First, outcomes from projects with different
Figure 1. Nutrition education can have widespread effects on
implementation protocols can be compiled into a report
health and independence. Reproduced with permission from
Hersey JC.12 that reflects the outcomes of FSNE nationally. The CNE
Journal of Nutrition Education and Behavior Volume 37 Number 4 July • August 2005 199

Logic Model promotes commonality among programs but completed CNE Logic Model was used for the development
leaves the nutrition educator free to select a behavior of a prototype Web-based system for data input and training.
change theory to drive development and implementation
of program content. Second, it was assumed that the “sys-
DECISIONS MADE IN THE DEVELOPMENT
tems, policy and environmental change” program frame-
OF THE CNE LOGIC MODEL
work could be applied to all 4 content areas to specify 3
levels of implementation: individuals and households; com- Dietary Quality and Physical Activity
munities and institutions; and social structure, policy,
and/or practices.This reflects an adaptation and simplifica- The annual Nutrition Education Plan Guidance recommends
tion of the socioecological model described by McLeroy et subject matter for nutrition education.The 2001 document14
al.13 Finally, the 4 content areas (dietary quality, food safety, contained a goal statement indicating that the main focuses
food security, and shopping behavior/food resource man- of nutrition education in FSNE are the Dietary Guidelines
agement) are assumed to be conceptually related but dis- for Americans15 and the Food Guide Pyramid.16 Therefore,
tinct components of a nutrition education process that specific skills and behaviors emphasized in this logic model
leads to personal health and nutritional independence.12 component would be those related to meeting recommended
The full CNE Logic Model can be viewed at http://www. servings and portion sizes of the Food Guide Pyramid;
uwex.edu/ces/lmcourse/. increased variety within each food group; increased intake of
Each subgroup was charged with responsibility for whole grains, certain nutrients, and fiber; and decreased
reviewing one of the white papers developed for FNS and intake of foods high in fat, sugars, and salt. In addition, skills
ERS5-9 and other current and relevant literature.The model’s and behaviors related to fitness, physical activity, and achiev-
components were peer-reviewed by FSNE project coordina- ing or maintaining a healthy weight were included. In con-
tors from the Extension/Land-Grant System and other agen- sideration of the increasing concern about obesity, its role in
cies. Model components were revised and incorporated into chronic diseases, and the importance of physical activity in
a single model with common language and appearance.The preventing obesity and chronic disease,17 the name of this

Figure 3. The Community Nutrition Education Logic Model. Reproduced with permission. Copyright 2004 by the Board of Regents of the
University of Wisconsin. Enhancing program performance with logic models. Available at: http://www.uwex.edu/ces/lmcourse/. Accessed
January 18, 2005.
200 Medeiros et al/COMMUNITY NUTRITION EDUCATION LOGIC MODEL

section was changed from “Dietary Quality” to “Dietary to meet the need for training. Enhancing Program Performance
Quality and Physical Activity.” with Logic Models (http://www.uwex.edu/ces/lmcourse/)
A number of assessment instruments previously tested for has 2 modules. Module 1,“Logic Model Basics,” provides an
reliability, validity, and sensitivity to change8 were adapted to overview of what constitutes a logic model, its basic com-
short-, medium-, and long-term indicators for the dietary ponents, and how they can be used to design and evaluate
quality/physical activity component. Consideration was educational programs. Module 2,“Introducing the Commu-
given to whether it was feasible for state and local nutrition nity Nutrition Education Logic Model,” defines the need for
educators to collect data on the indicators. a common framework to evaluate community nutrition
education. It includes the complete CNE Logic Model and
Food Safety outlines a protocol for administrative units to use in report-
ing their program efforts. Evaluation specialists and state
Practicing personal hygiene, cooking foods adequately, FSNE project coordinators from across the country reviewed
avoiding cross-contamination, keeping foods at safe temper- Enhancing Program Performance with Logic Models for content
atures, and avoiding food from unsafe sources were the con- and face validity, response burden, and potential usefulness,
structs suggested in the white paper as the organizational and it was refined based on feedback prior to it becoming
structure for planning and evaluating outcomes of food available on the Internet.
safety programs.7 Behaviors that have the greatest impact on
the incidence of foodborne illnesses were recommended as
INITIAL TESTING OF THE CNE LOGIC MODEL
the focus of program content.18 More recent studies have
documented priority behaviors that would be expected to Pilot Testing
have the greatest impact on disease incidence.19-21 The most
common behaviors that will impact the control of numerous In March 2003, FSNE project coordinators within the
pathogens are proper hand washing, use of thermometers to Cooperative Extension/Land-Grant University system and
verify safe and adequate cooking, washing surfaces and equip- other institutions were asked to complete a report using a
ment with hot, soapy water to prevent cross-contamination, pilot version of the CNE Logic Model template.23 Forty-
and avoiding foods that cannot be made safe by common three of 49 (87.7%) state or territory coordinators voluntar-
household preparation methods, such as raw sprouts.19 ily responded to this request.The intention of the pilot study
was to determine if national highlights could be captured
Food Security and Shopping Behavior/Food without losing the richness and flexibility found in nutrition
Resource Management education implemented according to locally identified needs
and resources.This procedure for compiling data from across
Food security and food resource management are highly the nation was consistent with a recommendation by Greg-
interrelated.22 The long-term outcomes of both concepts son et al to “…synthesize information from multiple sources
relate to having sufficient nutritious food in socially accept- to draw conclusions that are broad enough for generalization
able ways. Clear definitions that delineate the difference yet specific enough to be useful to federal, state, and com-
between food security and food resource management were munity stakeholders.”5(pS13)
not found in the literature. Food resource management over- States in the sample reported 5 214 654 direct contacts
laps with food security because minimally adequate through classes, workshops, one-on-one teaching, and group
resources must be available for education to extend food discussion. An additional 32 million indirect contacts were
resources and improve dietary quality.6 made through newsletters, public service announcements,
After much deliberation, the Committee determined that displays, or health fairs. Educational contacts were carried
for the CNE Logic Model, food security would be distin- out in cooperation with 13 835 state and local public and
guished by emergency and nonemergency food support private partners. Social marketing campaigns were con-
mechanisms. This means that education for individuals ducted in 18 states.This information provides an indication
would focus on accessing alternative or emergency food sys- of the national scope of FSNE.
tems and education of the community on the provision of Forty states and territories (93% of all reporting) submit-
emergency food systems. Food resource management is dis- ted 349 FSNE impact statements showing changes in diet
tinguished by what people can do with nonemergency food quality (44% of the examples), food safety (28%), food secu-
resources within personal, family, and social supports. rity (7%), and shopping behavior/food resource manage-
ment (7%).This indicates that the majority of programs are
focused on diet quality and that all components of the CNE
DEVELOPMENT OF Logic Model may not be equally implemented. The over-
WEB-BASED TRAINING MODULES whelming majority of impact statements (343 of 349) were
at the individual and household levels. The low number of
Training is needed to help educators use this complex new community or institutional and social policy impacts
model. A self-directed, Web-based approach was developed reported may have been due to a lack of focus in those areas,
Journal of Nutrition Education and Behavior Volume 37 Number 4 July • August 2005 201

less developed skills for recognizing impacts at the commu- research to refine and test the model and to develop related
nity or social policy levels, less knowledge on how to report training materials that lead to desired outcomes and goals.
impacts at those levels, or fewer instruments available for The final goal of community nutrition education is to
effectively measuring such impacts. achieve personal health and nutritional independence (see
Examples of impact statements provided by states were Figure 1). This can be better achieved by helping people
diverse because state plans are based on local needs, with learn how to make healthful food choices consistent with
resulting differences in programming efforts, methods, and the Dietary Guidelines for Americans15 and the Food Guide
measures of accountability. Many impact statements reflected Pyramid.16 The application of the CNE Logic Model to
short-term outcomes (gains in knowledge, skills, or intent to other nutrition education programs should allow for better
change) (46%) or medium-term behavioral impacts (52%) at documentation of the progress toward meeting that final
the individual and household level.The large percentage of goal and could detect the cumulative impact of program-
impacts reflecting behavioral change is not entirely surpris- ming for the first time.
ing considering recent emphasis in the literature and in
Nutrition Education Plan Guidance3 to teach for changed
behavior.Whether impacts reported actually reflect changed ACKNOWLEDGMENTS
behavior is beyond the scope of this work.24
Appreciation is extended to Jane Coulter (retired), Anna
Follow-up Mae Kobbe, Mary M. Gray, and Karen Konzelmann (retired)
at CSREES/USDA; Linda Kay Benning, National Associa-
A second work group was convened in February 2004 with tion of State Universities and Land-Grant Colleges, Exten-
9 members representing research, program, and network per- sion Directors/Administrators of the Cooperative Exten-
spectives in the Extension/Land-Grant University System sion/Land-Grant University System; Joanne Guthrie and
and in public health agencies, the two primary types of Eileen Stommes, ERS, USDA; Ellen Henert, University of
providers of FSNE. The workgroup was given a 3-fold Wisconsin; and Michael Newman and Stephen Green, Mis-
charge: review the 2002 state reports in depth and use that sissippi State University. We also wish to acknowledge staff
information to make recommendations for refinements to from the USDA FNS, who reviewed the CNE Logic Model
the logic model, including content and training needs; at various stages of development.
develop common definitions of terms in concert with other
partners; and provide input into the development of an on-
line reporting system to facilitate easier completion of the REFERENCES
reports and use for subsequent state and local planning. Ini-
tial findings of the committee, also supported by process of 1. National Center for Chronic Disease Prevention and Health Promo-
data review during compilation of the 2002 report, pointed tion, Centers for Disease Control and Prevention. Physical activity eval-
to the need for clearer definitions to ensure commonality in uation handbook. Available at: www.cdc.gov/nccdphp/dnpa/physical/
reporting, examples to aid the user, a pop-up menu of handbook/index.htm. Accessed January 18, 2005.
choices for the on-line reporting system; and training on 2. United Way of America. Achieving and measuring community out-
how to use the model in writing objectives, program devel- comes: challenges, issues, some approaches. Available at: http://
opment and planning, and aligning all program components. national.unitedway.org/outcomes/files/cmtyout1.pdf.Accessed January
18, 2005.
3. Food Stamp Program Accountability Division, US Dept of Agriculture.
IMPLICATIONS FOR Nutrition Education Plan Guidance, Fiscal Year 2005. Alexandria,Va: Food
RESEARCH AND PRACTICE and Nutrition Service; March 2004.
4. Contento IR, Balch GI, Bronner YL, et al. The effectiveness of nutri-
The use of the CNE Logic Model allows for flexibility across tion education policy, programs, and research: a review of research.
multiple nutrition education programs and was designed to J Nutr Educ. 1995;27:279-418.
have application beyond FSNE. Using this framework per- 5. Gregson J, Foerster SB, Orr R, et al. System, environmental, and policy
mits employment of a variety of curricula and evaluation changes: using the social-ecological model as a framework for evaluat-
tools and allows data to be aggregated by programs or as ing nutrition education and social marketing programs with low-
summaries of state nutrition education programming. Units income audiences. J Nutr Educ. 2001;33(Suppl 1):S4-S15.
are able to conduct programming and social marketing cam- 6. Hershey JC,Anliker J, Miller C, et al. Food shopping practices are asso-
paigns that are appropriate to their particular audience and ciated with dietary quality in low-income households. J Nutr Educ.
needs assessments and report in a way that allows for data to 2001;33(Suppl 1):S16-S26.
be put into a broader perspective. 7. Medeiros L, Hillers V, Kendall P, Mason A. Evaluation of food safety
This report represents a starting point for testing the fea- education for consumers. J Nutr Educ. 2001;33(Suppl 1):S27-S34.
sibility and application of the CNE Logic Model, which 8. McClelland JW, Keenan DP, Lewis J, et al. Review of evaluation tools
others will be able to build on.There is a need for additional used to assess the impact of nutrition education on dietary intake and
202 Medeiros et al/COMMUNITY NUTRITION EDUCATION LOGIC MODEL

quality, weight management practices, and physical activity of low- 17. Rippe JM,Aronne LJ, Coulson AM, et al.The obesity epidemic: a man-
income audiences. J Nutr Educ. 2001;33(Suppl 1):S35-S48. date for a multidisciplinary approach. J Am Diet Assoc. 1998;98(10
9. Keenan DP, Olson C, Hersey JC, Parmer SM. Measures of food insecu- Suppl 2):S55-S61.
rity/security. J Nutr Educ. 2001;33(Suppl 1):S49-S58. 18. Medeiros LC, Hillers VN, Kendall PA, Mason A. Food safety education:
10. Alter C, Egan M. Logic modeling: a tool for teaching critical thinking what should we be teaching? J Nutr Educ. 2001;33:108-113.
in social work practice. J Soc Work Educ. 1997;33:85-102. 19. Medeiros LC, Kendall P, Hillers V, Chen G, DiMascola S. Identification
11. Alter C, Murty S. Logic modeling: a tool for teaching practice evalua- and classification of consumer food handling behaviors for food safety
tion. J Soc Work Educ. 1997;33:103-118. education. J Am Diet Assoc. 2001;101:1326-1332, 1337-1339.
12. Hersey JC. Integrating evaluation tools to assess nutrition education. 20. Hillers VN, Medeiros LC, Kendall P, Chen G, DiMascola S. Consumer
J Nutr Educ. 2001;33(Suppl 1):S2-S3. food handling behaviors associated with prevention of thirteen food-
13. McLeroy KR, Bibeau D, Steckler A, Glanz K.An ecological perspective borne illnesses. J Food Protection. 2003;66:1893-1899.
on health promotion programs. Health Educ Q. 1988;15:351-377. 21. Kendall P, Medeiros LC, Hillers V, Chen G, DiMascola S. Food handling
14. Food Stamp Program Accountability Division, US Dept of Agriculture. behaviors of special importance for the pregnant, young, elderly and
Nutrition Education Plan Guidance, Fiscal Year 2001. Alexandria,Va: Food immune compromised. J Am Diet Assoc. 2003:103:1646-1649.
and Nutrition Service; March 2000. 22. Weimer J, McKinney P, Benning LK.Tools to assess nutrition education
15. US Dept of Agriculture and Health and Human Services. Nutrition and in low-income families. J Nutr Educ. 2001;33(Suppl 1):S1.
Your Health: Dietary Guidelines for Americans 2005.Available at: http://www. 23. Little DM, Newman ME. Food Stamp Nutrition Education within the
healthierus.gov/dietary guidelines/.Accessed January 18,2005. Cooperatuive Extension/Land-Grant University System – National
16. Center for Nutrition Policy and Promotion, US Dept of Agriculture. report FY2002.Available at: www.csrees-fsnep.org.Accessed January 18,
The Food Guide Pyramid. Washington, DC: US Government Printing 2005.
Office; 1992. Home and Garden Bulletin, No. 252. Available at: 24. Byers T. On the hazards of seeing the world through intervention-
http://www.usda.gov/cnpp/pyramid.html. Accessed April 5, 2004. colored glasses. Am J Clin Nutr. 2003;78:904-905.