Journal of Food Distribution Research

Volume 42, Issue 3

Self Efficacy as a Mediator of the Relationship between

Dietary Knowledge and Behavior
Arbindra Rimala, Wanki Moonb, Siva K. Balasubramanianc and Dragan Miljkovicd
a
Professor, Agribusiness-Department of Agriculture, Missouri State University,
901 S. National Avenue, Springfield, Missouri, 65810, U.S.A.
b
Associate Professor, Department of Agribusiness Economics, Southern Illinois University, MC4410,
Carbondale, Illinois, 62901, U.S.A.
c
Harold L. Stuart, Professor of Marketing, Stuart School of Business, Illinois School of Technology,
3424 S. State Street, Chicago, Illinois, 60616, U.S.A.
d
Professor, Agribusiness and Applied Economics, North Dakota State University, 614A Barry Hall,
Fargo, North Dakota, 58108, U.S.A.

Abstract
Translating the dietary knowledge among individuals into healthy behavior remains a
challenging task. This study examines the causal relationship between dietary knowledge and
behavior by including self-efficacy in the models.

A series of regression models were developed based on Baron and Kenny (1986) to assess
whether self-efficacy mediated the link between the predictor variables and dietary behavior.
Regression analyses supported the hypothesized relationships that self-efficacy mediates effects
of dietary knowledge and social influences on dietary behavior. Self-efficacy also accounted for
variance in eating behavior not explained by knowledge or demographic variables.

Keywords: dietary knowledge, dietary behavior, self-efficacy


Corresponding author: Tel: +1 417.836.5094
Email: arbindrarimal@[Link]
W. Moon: wmoon@[Link]
S.K. Balasubramanian: siva@[Link]
D. Miljkovic: [Link]@[Link]

28
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

Introduction and Objectives

Increased availability of nutritional information has been successful in enhancing public
awareness of the importance of healthy diet and lifestyles. The important issue is whether
enhanced nutrition and health awareness has any significant impact on consumers’ actual dietary
behavior. The data from the healthy eating index (HEI) show that although dietary quality has
improved over the past years, the diets of most Americans need improvements in several aspects
(Kennedy et al. 1999; Guo et al. 2004). Previous studies have examined the influence of health
behavior through informational campaigns, followed by the expected change in the attitude and
desired behavioral changes in areas like smoking, obesity, and HIV/AIDS (Perry et al. 1980;
Stern et al. 1982; Nwokocha and Nwakoby 2002.) While the above studies have reported mixed
results of success, studies evaluating the relationship between nutrition knowledge and dietary
behavior have found no direct correlation between the two (Putler and Frazao 1994; Sapp 1991).

Clearly, the evidence from the above studies suggests that the impact of additional information
and knowledge on actual consumer behavior is an empirical issue. This is in contrast to the
premises of the rational choice theory which is the basis for traditional neoclassical theory of
demand and consumer choice (e.g., Mas-Colell, Whinston and Green 1995). The implausibility
of the rational choice axioms has been documented by many economists including, among
others, the Nobel Prize Laureate Kahneman (1994), or more recently Miljkovic (2005).
Therefore, translating the dietary knowledge among individuals into healthy behavior remains a
challenging task for economic modelers, and in turn the food and health policy makers. Relying
on behavioral sciences theories such as the social cognitive theory (SCT), the objective of this
study is to examine the causal relationship between dietary knowledge and behavior by including
self-efficacy in the models.

Self-efficacy is defined as a person’s ability of exerting self-control in changing his/her behavior.

Hence, the objective of this study may be more specifically stated as to empirically address the
question of whether the predictor variables such as dietary knowledge affect only self-efficacy,
or dietary behavior, or both. The self-efficacy component of the SCT has been widely used by
many researchers to explain human behavior with regard to, for example, phobias (Bandura
1983) , smoking (Schinke et al. 1985), drug use (Hays and Ellickson 1990), addiction (Marlatt
Baer, and Quigley) and food choices (Parcel et al. 1995; Steptoe, et al. 1995). Researchers have
suggested that self-belief that includes self-efficacy plays a mediating role in relation to
cognitive activities. Bandura (1997) explained self-efficacy belief as “beliefs in one’s capability
to organize and execute the courses of action required to manage prospective situation.” A large
amount of previous research has generally supported the basic notion proposed by Bandura
(1986 and 1997) that efficacy beliefs mediate the effects of skills on performance by influencing
effort, persistence and perseverance (Schunk 1991; Bouffard-Bouchard 1990; and Schunk and
Hanson 1985). Corwin et al. (1999) reported that many components from SCT including self-
efficacy had significant correlation with the diet related behavior of children. In a study among
fourth graders, she reported that the mean dietary exposures scores for low-fat food selection
was significantly higher for those children who had scored highest levels of confidence about
lower fat food choices than those with lower levels of confidence.
29
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

Another study designed to examine the social-cognitive determinants of health behaviors

including physical exercise, smoking, alcohol consumption, and preventive nutrition (Schwarzer
and Renner 2000) distinguished between action self-efficacy (preintention) and coping self-
efficacy (postintention) as two phases of optimistic self- beliefs. The study reported that the
importance of perceived self-efficacy increased with the age of the respondent and their body
weight.

A person’s health related self efficacy is influenced by his/her health knowledge and other socio-
demographic background information. Since self-efficacy itself is explained by the dietary
knowledge of individuals (Slater 1989), it is likely to play a mediating role in the relationship
between healthy behaviors and dietary knowledge. Consumers with higher levels of self-
efficacy are more likely to sustain a healthy behavior with regard to food choices compared to
those with lowers level of self-efficacy.

Theoretical and Empirical Models

The preceding discussion points to a causal flow from dietary knowledge (hereafter, we call
these predictor variables) and socio-demographic characteristics to self-efficacy and/or dietary
behavior. At this point, an empirical question that remains to be determined is whether the
predictor variables affect only self-efficacy, or dietary behavior, or both. We propose a
mediation model here. More specifically, we hypothesize that (a) the predictor and socio-
demographic variables influence both self-efficacy and dietary behavior, and (b) these variables
influence dietary behavior primarily via their link to dietary knowledge. For example, when
consumers possess a high level of dietary knowledge, they are predisposed to exert a greater
control over their diets and lifestyle, thereby adopting a healthy dietary behavior.

The hypotheses above underscore the notion of mediation. In other words, the mediation
approach recognizes that consumers’ self-control (efficacy) over diet and lifestyle can mediate
the effects of the predictor variables (dietary knowledge) on the dietary behavior (Baron and
Kenny 1986). Figure 1 (as adapted from Baron and Kenny 1986) illustrates this modeling
approach using self-efficacy as mediators of the relationship between dietary behavior and
predictor variables. The figure depicts three causal paths in a model of how overall dietary
behavior is formed: (i) the direct impact of the predictors on dietary behavior (path a); (ii) the
path from the predictors to the mediators (path b); and (ii) the impact of mediators on dietary
behavior (path c).

In this study, the mediating hypothesis is tested using the following four criteria adopted from
Judd and Kenny (1981) and Baron and Kenny (1986): a) the self-efficacy of individuals
(mediator) has a statistically significant impact on dietary behavior; b) dietary knowledge and
socio-demographic variables (predictors) have significant influence on dietary behaviors; c)
dietary knowledge exerts a significant influence on diet related self-efficacy of individuals; and
d) the effects of dietary knowledge is either diminished or no longer significant when self-
efficacy is controlled for the dietary behavior equations.

30
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

Figure 1. Conceptual model depicting the mediating role of self-efficacy between dietary
behavior and predictor variables (adapted from Baron and Kenny, 1986).

Following Baron and Kenny (1986) and Judd and Kenny (1981), a series of regression models
were developed to assess whether self-efficacy mediated the link between the predictor variables
and dietary behavior:

Model 1: BEHAVIOR = b10 + b11 DIETARY KNOWLEDGE + e

Model 2: BEHAVIOR = b20 + b21 DIETARY KNOWLEDGE +b22 FFICACY + e

Model 3: BEHAVIOR = b30 + b31 DIETARY KNOWLEDGE + b32 FFICACY +b33 AGE
+ b34 GENDER + b35 INCOME+ b36 EDUC+ b37 RACE+ b38 HOUSEHOLD SIZE + e

Comparing estimated coefficients across Models 1 - 3 allows us to assess whether self-efficacy

mediates the effects of the predictor variables on dietary behavior. To illustrate, assume that
dietary knowledge exerts a statistically significant influence on behavior in Model 1. If dietary
knowledge in the Model 2 has a negligible effect on behavior, it indicates that the effect of
dietary knowledge is largely transmitted via the degree of self-control consumers can exercise on
their diet and lifestyle. Second, if the effect of self-efficacy in Model 3 differs little from that in
Model 2, it suggests that impacts of efficacy on diet behavior remain stable despite the presence
of other predictors (socio-economic profile) in the model. The last case is a combination of the
previous two: although the effects of efficacy in Model 3 are smaller than those in Model 2, they

31
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

remain statistically significant. This indicates that the effects of dietary knowledge are partially
mediated by efficacy.
The empirical model posits that a participant’s dietary behavior is a function of dietary
knowledge, self control (efficacy) in changing health behavior with regard to food choices and
life-style and various socio-economic characteristics of individuals. We are interested in
explaining consumption intensity with regard to fruits, vegetables, and nutrients such as
cholesterol and fat rather than number of times someone consumed them in the past. The model,
therefore, can be formally written as:

Uj = β’Zj + Ɛj

Where Uj is a participant’s actual dietary behavior and Zj is a vector of explanatory variables

including participant’s socio-economic profile. Although Uj is unobserved, what is observed is
the expressed intensity of consumption represented by the rank-ordered dependent variables, R,
where:
R = 0 if Uj ≤ 0
R= 1 if 0 < Uj ≤ µ1
R =2 if µ1< Uj ≤ µ2
R = w if µw-2 < Uj
where the µ’s are the threshold variables or cutoff points that provide the ranking of intensity in
consuming specific dietary item. The lowest ranked outcome, R=0 represents the situation when
a statement (e.g. I eat a lot of) regarding a specific dietary item does not represent a participant at
all. Highest ranked outcome, R=w, represents the situation when a statement represents
“extremely well.”

The dependent variable in the models were measured using ordinal measures (1,2,3, 4 and 5.)
Hence, an ordered probit model (Long 1997; Greene 1993) was used to conduct the regression
analysis. Value of 1 indicated when a statement regarding a dietary item (e.g. I eat a lot of fresh
fruits) did not describe a participant “at all”; value of 2 indicated that it described “slightly”;
value of 3 indicated that it described “somewhat”; value of 4 indicated that it described “very
well” and value of 5 indicated that it described “extremely well.”

The Data
In the summer of 2007, a national survey among United States household was conducted. The
survey was administered online by Ipsos-Observer, a private consulting firm specializing in
consumer research and public opinion poll on socially important issue including tracking trends
in food consumption. This firm maintains an on-line panel that consists of 400,000 households.
Approximately stratified by geographic regions, income, education, and age to correspond to the
2000 US census, a sample of 9000 households were drawn out of the online panel in a manner
that is representative of the US population. A total of 3,456 households completed the surveys,
resulting in a 38.4% response rate. Sample households were sent e-mails soliciting information
regarding their food consumption behavior and household characteristics. Each e-mail included
a unique URL (keyed to the respondent’s ID) to direct the respondent to the survey website. In
32
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

addition to socio-economic characteristics of sample households, survey instruments included

questions relating to three key components in the mediating model: dietary knowledge, dietary
behavior and diet related self-efficacy.
Respondents were asked dietary behavior questions about fresh fruits, fresh vegetables, fat and
cholesterol (Table 1). They were asked to respond as to how well the statements described their
dietary behavior using a scale of one to five where one represented “not at all” and five
represented “extremely well.” Four statements to measure diet related self-efficacy were read to
the participants in the survey (Table 2). The respondents were asked “How likely are you to read
nutritional labels on food packages carefully”, “How likely are you to change diet to reduce the
risk of certain diseases”, “How likely are you to exercise at least three times per week” and
“How likely are you to prevent health problems before feeling any symptoms” Respondents’
reported self-efficacy were recorded on a 5-point scale. All responses were first coded such that
the higher values represented high level of self-efficacy. Respondents were asked to respond as
to how well the statements described the self-control (efficacy) in changing health behavior with
regard to food choices and life-style. The lowest degree of self-control was represented by the
response “extremely unlikely” and the highest degree of self control was represented by the
response “extremely likely.” The percentage of respondents who reported each level of self-
control were reported in Table 2. A test was conducted to evaluate the internal consistency of the
four statements. The computed test statistic showed that the four statements had a high level of
consistency (Cronbach’s α = 0.84) in measuring levels of self-efficacy. A composite self-
efficacy index was created by summing up the reported scores for each statement and dividing
by four. The higher the index value the higher the overall level of self control.

Table 1. Food Consumption Behavior of US households (n=3056).

How well each of I eat a lot of I eat a lot of I am actively trying I am actively trying
the statements fresh fruits fresh vegetables to consume less fat to consume less
describes you? in my diet cholesterol in my diet
1 = Not at all 5.9% 5.5% 8.1% 12.2%
2 = Slightly 19.8% 17.0% 13.4% 16.1%
3 = Somewhat 33.8% 33.0% 31.8% 31.0%
4 = Very well 25.9% 29.2% 31.8% 26.7%
5 = Extremely well 14.5% 15.2% 14.9% 13.9%

Table 2. Reported level of self-control (Efficacy) in changing health behavior with regard to
food choices and life-style (n=3056).
Percentage of Respondents
1= 2= 3= 4= 5=
How likely are you to: Extremely Unlikely Slightly Somewhat Very much Extremely Likely
Read nutritional labels on
food packages very carefully? 12.5% 19.5% 27.8% 24.6% 15.5%
Change diet to reduce the risk
of certain diseases? 23.3% 18.9% 28.9% 20.1% 8.7%
Exercise at least three times
per week? 25.9% 20.2% 19.6% 17.3% 16.9%
Prevent health problems
before feeling any symptoms 9.0% 17.9% 35.2% 27.5% 10.4%
Note. Cronbach’ s consistency test (α) was 0.85
33
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

A knowledge of the diet health relationship was measured using an instrument similar to the one
used by Moorman and Matulich, (1993), who defined health knowledge as the extent to which
consumers have enduring health-related cognitive structures Respondents were asked to link or
match each of the eleven nutrients (i.e., sodium, calcium, vitamin A, protein, vitamin C, iron,
vitamin D, carbohydrates, saturated fat, potassium, and dietary fiber) with an appropriate health
consequence from a list: high blood pressure, strong bones, healthy eyes, amino acids, anticancer
power, oxygen, absorb calcium, conversion to sugar and fueling the body, cardiovascular
disease, and balancing sodium. An index of dietary knowledge was constructed by adding all
correct answers for each respondent. Hence, the index ranges from a minimum of 0 (representing
no dietary knowledge) to a maximum of 11 (representing highest dietary knowledge.) The mean
dietary knowledge score was 6.09 (Table 3) which means an average respondent could provide
six correct matches out of eleven.

Table 3 reports descriptive statistics for other (socio-economic) explanatory variables -including
gender, age, household income, education level of the respondent, household size and ethnic
background. Over 50% of the respondents were female. The average age of the respondent was
50 years. Household income was reported in income groups represented by numerical values.
For example, 1 represented less than $5,000 and 25 represented more than $250,000. In the
analysis, mid-points in each income group were used to obtain household income in dollars. The
average household income among the sample respondents was $67,377. Average household size
was 2.6 members. Nearly three fourths of respondents were white.

Table 3: Description of other explanatory variables used in the analysis.

VARIABLES DESCRIPTION Mean Std.
Deviation
Dietary Knowledge Total number of dietary questions answered correctly 6.085 3.142
(0 to 11).
Socio demographics
Gender 1 = female; 0 = male 0.501 0.500
Age Respondents’ age in years 49.722 14.754
Income 1 = less than $5,000; 25 = $250,000 or more $67,377 $38,292
Education 1 = college or more than college education; 0= 0.649 0.477
otherwise
Household Size Number of household member 2.612 1.399
Ethnic background 1 if white; 0 otherwise 0.734 0.442

A Pearson correlation matrix including all the independent variables was generated to examin
any potential multicollinearity in the regression models. While many coefficients were
statistically significant at 0.05 level, the size of the coefficient was very small. The largest
coefficient was 0.21. Hence, it was determined that multicollinearity was unlikely in the
proposed regression models
.
Results and Implications
Ordered probit models for each of the four dietary behaviors: fresh fruits, fresh vegetables, fat
and cholesterol were run and reported in Tables 4 to 7 (see Appendix). For all models the null
34
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

hypotheses that all parameters were jointly equal to zero were rejected using χ2 statistics at the
0.01 significance level. Based on the collinearity diagnostics (Belsley et al., 1980), no
collinearity problems were detected in the analyses. Marginal effects of the independent
variables were also estimated but not reported due to the space consideration. Initially, only
knowledge was used as the explanatory variable. Self-efficacy and socio-demographic variables
were added in subsequent runs.

The coefficients for the relationship between dietary behavior and knowledge are positive and
significant, as one may have expected, in Model 1 of the all four dietary behaviors. This result
suggests only that more dietary knowledge translates into more responsible and healthy dietary
behavior, but it does not explain or clarify the mechanism or the process which leads more
dietary knowledge to transfer into more responsible and healthy dietary behavior. This aspect of
the problem is explained in Models 2 and 3.

In Model 2, when the influence of self-efficacy was added, the impact of dietary knowledge
decreased but remained statistically significant for vegetables and fat while it became statistically
insignificant for fruits and cholesterol. The coefficients measuring the impact of the self-efficacy
on dietary behavior are all positive, statistically significant, and much larger in size than the
coefficients associated with the knowledge variable. The pseudo R-squared for each of the four
dietary items increased by a huge magnitude when self-efficacy was added to the models. All
the above results from the regression analysis of Model 2 supported the hypothesized
relationships that self-efficacy mediates effects of dietary knowledge and social influences on
dietary behavior for each of the four dietary items.

Self-efficacy also accounted for variance in eating behavior not explained by knowledge or
demographic variables. However, the effect of self-efficacy on dietary behavior in Model 3,
albeit remaining statistically significant, decreased substantially in the cases of both fat and
cholesterol. Moreover, the pseudo R-squared in these two regressions decreased when
demographic variables were added. While the impact of all demographic variables on the dietary
habits in fruit consumption behavior equation is statistically significant, and the impact of all
demographic variables but the education is statistically significant in the vegetables consumption
behavior equation, the demographic variables had almost no impact on consumption of fat
(except age) and cholesterol (except the household size).

The above results indicate that self-efficacy is the most important mechanism in impacting fat
and cholesterol consumption, while it is only one of the factors impacting the consumption of
both fruits and vegetables. This should come as no surprise: healthy nutrition implies eating
more of fruits and vegetables for most people while cutting out the consumption of fat and
cholesterol. Consuming more of anything is hardly considered a sacrifice while consuming less
of something often demands a great deal of self-control and discipline. The results in this study
are consistent with results in other studies which show that dieting, weight control and
preventive nutrition can be governed by self-efficacy beliefs. In intervention programs, clients
with higher level of self-control were less likely to relapse into their previous habit than those
with lower level of self-control (Chambliss and Murray 1979; Furhrmann and Kuhl 1998;

35
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

Schnoll and Zimmerman 2001; Long and Stevens. 2004; Luszczynska et al. 2007). Yet there is
no clear unique solution as for what the means to inducing dietary self-efficacy may be. For
example, some studies suggest that goal setting is the most critical way to induce self-efficacy in
dietary behavior (e.g., Robinson 1999; Baldwin and Galciglia 1997). Other studies suggest that
goal setting and self-monitoring combined increase the self-efficacy scores significantly (e.g.,
Schnoll and Zimmerman 2001). Also, other aspects of self-regulation and behavioral training
such as problem identification, problem solving, self-evaluation, or reinforcement may be critical
in inducing dietary self-efficacy (Hardeman et al. 2000). Hence, interventions and health
promotion campaigns should seek to directly address factors influencing diet related self-efficacy
instead of focusing on disseminating information only. In practice, for example, we often see
healthy foods such as fruits and vegetables being introduced on the menus of school and college
cafeterias or in restaurants. The availability of healthy foods coupled with self-efficacy driven
by dietary knowledge is likely to lead to an increased consumption of healthy foods. At the same
time, most restaurants and cafeterias sell foods rich in fat and cholesterol alongside the healthy
foods. Also, while often the consumers are aware of the negative impact fat and cholesterol may
have on their health due to numerous educational activities by health and nutrition professionals,
the low cost of that food coupled with the sugar enhanced, taste improving additives proves to be
irresistible to the average consumer (Miljkovic, Nganje, and de Chastenet 2008). It has been
shown that sweetened foods, i.e., an increased consumption of sugar, leads first to sugar
addiction and second to carbohydrate addiction and increased consumption of fats (Miljkovic
and Nganje 2008). Hence unavailability of unhealthy food or its availability at higher cost due to
“fat tax,” especially to children and adolescents who develop taste for unhealthy foods at an early
age, seems to be a reasonable pro-active approach to influence diet related self-efficacy.

References
Baldwin, T.T., and G.A. Galciglia. 1997. “Application of cognitive behavioral theories to dietary
changes in clients.” Health Education & Behavior 24: 357-368.

Bandura, A. 1983. “Self-efficacy determinants of anticipated fears and calamities.” Journal of

Personality and Social Psychology 45: 464-469.

Bandura, A. 1986. Social Foundations of Thought and Action: A Social Cognitive Theory.
Englewood Cliffs, NJ: Prentice Hall.

Bandura A. 1997. Self-efficacy: The exercise of control. New York: Freeman

Baron R. M. and D. A. Kenny. 1986. “The Moderator-Mediator Variable Distinction in Social

Psychological Research: Conceptual, Strategic, and Statistical Considerations.” Journal
of Personality and Social Psychology 51(6): 1173-1182.

Belsey, D.A., E. Kuh, and [Link]. 1980. Regression diagnostic, identifying influential data
and source of colliniearity. Wiley: New York.
36
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

Bouffard-Bouchard, T. 1990. “Influence of Self-Efficacy on Performance in a Cognitive Task”

Journal of Social Psychology 130:353-363.

Corwin, S.J.,R.G. Sargent, C.E. Rheaume, and R.P. Saunders. 1999. “Dietary Behavior among
Fourth Graders: A Social Cognitive Theory Study Approach” American Journal of
Health Behavior. 23: 182-197.

Chambliss, C.A. and E.J. Murray. 1979. “Efficacy attribution, locus of control, and weight loss”
Cognitive Therapy and Research 3: 349-353.

Fuhrmann, A. and J. Kuhl. 1998. “Maintaining a healthy diet: effects of personality and self-
reward versus self-punishment on commitment to an enactment of self-chosen and
assigned goals” Psychology and Health 13:651-686

Greene, W. 1993. Econometric Analysis. McMillan Publishing: New York.

Guo, X., B.A. Warden, S. Paeratakul and G.A. Bray. 2004. “Healthy eating index and obesity”
European Journal of Clinical Nutrition 58(12): 1580-1586.

Hardeman, W., S. Griffin, M. Johnston, A.L. Kinmonth, and N.J. Warcham. 2000. “Interventions
to prevent weight gain: a systematic review of psychological models and behaviour
change methods.” International Journal of Obesity 24: 131-143.

Hays, R.D., and P.L. Ellickson. 1990. “How generalizable are adolescents' beliefs about pro-drug
pressures and resistance self-efficacy?” Journal of Applied Social Psychology 20: 321-
340.

Judd, C.M. and Kenny, D.A. 1981. Process Analysis: Estimating mediation in Evaluation
Research. Evaluation Research 5: 602-619.

Kahneman, D., 1994. New challenges to the rationality assumption. Journal of Institutional and
Theoretical Economics 150 (1), 18–36.

Kennedy E., S.A. Bowman, M. Lino, S.A. Gerrior, and P.P. Basiotis 1999. “Diet quality of
Americans,” in America’s Eating Habits. Elizabeth Frazao eds., USDA, ERS
Agricultural Information Bulletin number 750.

Long, JoAnn D. and K. Stevens 2004. “Using Technology to Promote Self-Efficacy for Healthy
Eating in Adolescents.” Journal of Nursing Scholarship 26(2):134-139.

Long, J.S. 1997. “Regression Models for Categorical and Limited dependent Variables.” In
Advanced Quantitative Techniques in the Social Science Series (7): 114-147. Sage
Publication :Thousand Oaks, CA.

37
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

Luszczynska, Aleksandra, M. Tryburcy, and R. Schwarzer 2007. “Improving Fruit and

Vegetable Consumption: A Self-Efficacy Intervention Compared with a Combined Self-
Efficacy and Planning Intervention” Health Education Research 22(5):630-638.
Marlatt, G. A., J.S. Baer, and L.A. Quigley. 1995. “Self-efficacy and addictive behavior. In A.
Bandur (Ed.), Self-efficacy in Changing Societies (pp. 289-315). Cambridge University
Press: New York.

Mas-Colell, A., M. D. Whinston, and J.R. Green. 1995. Microeconomic Theory. Oxford
University Press: New York.

Miljkovic, D. 2005. “Rational choice and irrational individuals or simply an irrational theory: A
critical review of the hypothesis of perfect rationality.” Journal of Socio-Economics,
34(5): 623-636.

Miljkovic, D., W. Nganje and H. de Chastenet. 2008. “Economic factors affecting the increase in
obesity in the United States: the differential response to price.” Food Policy 33(1): 48-60.

Miljkovic, D., and W. Nganje. 2008. “Regional obesity determinants in the United States: a
model of myopic addictive behavior in food consumption.” Agricultural Economics 38:
375-384.

Moorman, C and E. Matulich 1993. “A model of consumers’ preventive health behaviors: The
role of health motivation and health ability. Journal of Consumer Research 20: 209-229.

Nwokocha, A.R. and B.A. Nwakoby. 2002. “Knowledge, Attitude, and Behavior of Secondary
(High) School Students Concerning HIV/AIDS in Enugu, Nigeria, in the Year 2000.”
Journal of Pediatric Adolescent Gynecology 15:93-6.

Perry, C., J. Killen, M. Telch, L.A. Slinkard, and B.G. Danaher. 1980. “Modifying Smoking
Behavior of Teenagers: A School Based Intervention” American Journal of Public
Health 70: 722-725.

Putler, D.S., and E. Frazao 1994. “Consumer Awareness of Diet-Disease Relationships and
Dietary Behavior: The Case of Dietary Fat,” Journal of Agricultural Economics Research
45:3-17.

Robinson, T.N. 1999. “Behavioral treatment of childhood and adolescent obesity.” International
Journal of Obesity 28(supplement): S52-S57.

Schnoll, Roseanne and B. Zimmerman 2001. “Self-Regulation Training Enhances Dietary Self-
Efficacy and Dietary Fiber Consumption.” Journal of American Dietetic Association
101(9): 1006-1011

38
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

Sapp, S.G. 1991. “Impact of Nutrition Knowledge within Expanded Rational Expectations
Model of Beef Consumption,” Journal of Nutrition Education 23:214-222.

Slater, M.D. 1989. “Social Influences and Cognitive Control as Predictors of Self-efficacy and
Eating Behavior” Cognitive Therapy and Research 13(3):231-245.

Steptoe, A., R. Sanderman and [Link]. 1995. “Stability and Changes in Health Behaviors in
Young Adults over a One Year Period.” Psychology of Health 10:155-167.

Schinke, S. P., L. D. Gilchrist and W. H. Snow 1985. “Skills intervention to prevent cigarette
smoking among adolescents.” American Journal of Public Health 75 (6): 665-667.

Schnoll, R., and B.J. Zimmerman. 2001. “Self-regulation training enhances dietary self-efficacy
and dietary fiber consumption.” Journal of the American Dietetic Association 101(9):
1006-1011.

Schwarzer, R. and B. Renner. 2000. “Social-Cognitive Predictors of Health Behavior: Action

Self-Efficacy and Coping Self-Efficacy” Health Psychology 19:487-495.

Shunk, D.H. 1991. “Self-efficacy and Academic Motivation.” Educational Psychologist 26: 207-
231.

Shunk, D.H. and A.R. Hanson. 1985. “Peer Models: Influence on Children’s Self-Efficacy and
Achievement.” Journal of Educational Psychology 77:313-322.

Stern, M.P., J. Pugh, S. Gaskill, and H. Hazuda. 1982. “Knowledge, Attitudes and Behavior
Related to Obesity and Dieting in Mexican Americans and Anglos: The San Antonio
Heart Study” American Journal of Epidemiology 115:917-928.

39
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

Appendix
Table 4. Mediation by efficacy in the relationship between dietary knowledge and fruit
consumption behavior: An Ordered Probit Model
Model1 Model2 Model3
Variables Coeff. P-value Coeff. P-value Coeff. P-value
ONE 1.399 0.000 0.265 0.000 -0.154 0.142
KNOW 0.028 0.000 0.002 0.706 -0.006 0.317
EFFICACY 0.563 0.000 0.551 0.000
AGE 0.005 0.000
GENDER 0.215 0.000
INCOME 0.001 0.005
EDUCA 0.090 0.036
RACE -0.125 0.003
HHSIZE 0.043 0.001
Mu( 1) 0.914 0.000 1.012 0.000 1.025 0.000
Mu( 2) 1.812 0.000 2.006 0.000 2.031 0.000
Mu( 3) 2.628 0.000 2.910 0.000 2.945 0.000
*
Pseudo-R-Squared 0.01 0.32 0.35
*
R2ML = 1 – exp(-G2/N), where G2 = -2 ln [L(Mα)/L(Mβ)]; Mα = restricted likelihood, Mβ = Unrestricted Likelihood,
and N=Number of observation (Maddala. 1983).

Table 5. Mediation by efficacy in the relationship between dietary knowledge and vegetable
consumption behavior: An Ordered Probit Model
Model1 Model2 Model3
Variables Coeff. P-value Coeff. P-value Coeff. P-value
ONE 1.361 0.000 0.171 0.008 -0.415 0.000
KNOW 0.041 0.000 0.016 0.005 0.007 0.286
EFFICACY 0.592 0.000 0.579 0.000
AGE 0.008 0.000
GENDER 0.246 0.000
INCOME 0.002 0.000
EDUCA 0.050 0.243
RACE -0.115 0.007
HHSIZE 0.041 0.002
Mu( 1) 0.853 0.000 0.955 0.000 0.976 0.000
Mu( 2) 1.756 0.000 1.962 0.000 2.002 0.000
Mu( 3) 2.649 0.000 2.960 0.000 3.016 0.000
*
Pseudo-R-Squared 0.03 0.36 0.39
*
R2ML = 1 – exp(-G2/N), where G2 = -2 ln [L(Mα)/L(Mβ)]; Mα = restricted likelihood, Mβ = Unrestricted Likelihood,
and N=Number of observation (Maddala. 1983).

40
November 2011 Volume 42, Issue 3
Rimal et al. Journal of Food Distribution Research

Table 6. Mediation by efficacy in the relationship between dietary knowledge and fat
consumption behavior: An Ordered Probit Model
Model1 Model2 Model3
Variables Coeff. P-value Coeff. P-value Coeff. P-value
ONE 1.087 0.000 -0.617 0.000 0.524 0.000
KNOW 0.055 0.000 0.019 0.002 0.025 0.000
EFFICACY 0.889 0.000 0.128 0.000
AGE 0.006 0.000
GENDER 0.034 0.368
INCOME 0.001 0.125
EDUCA 0.012 0.773
RACE -0.003 0.942
HHSIZE 0.006 0.656
Mu( 1) 0.621 0.000 0.782 0.000 0.607 0.000
Mu( 2) 1.507 0.000 1.902 0.000 1.481 0.000
Mu( 3) 2.475 0.000 3.120 0.000 2.439 0.000
Pseudo-R-Squared* 0.05 0.60 0.26
*
R2ML = 1 – exp(-G2/N), where G2 = -2 ln [L(Mα)/L(Mβ)]; Mα = restricted likelihood, Mβ = Unrestricted Likelihood,
and N=Number of observation (Maddala. 1983).

Table 7. Mediation by efficacy in the relationship between dietary knowledge and cholesterol
consumption behavior: An Ordered Probit Model
Model1 Model2 Model3
Variables Coeff. P-value Coeff. P-value Coeff. P-value
ONE 0.932 0.000 -0.662 0.000 0.384 0.000
KNOW 0.040 0.000 -0.009 0.147 0.004 0.542
EFFICACY 0.846 0.000 0.333 0.000
AGE 0.002 0.077
GENDER -0.023 0.556
INCOME 0.000 0.382
EDUCA -0.012 0.773
RACE -0.054 0.201
HHSIZE -0.038 0.004
Mu( 1) 0.598 0.000 0.748 0.000 0.590 0.000
Mu( 2) 1.413 0.000 1.768 0.000 1.416 0.000
Mu( 3) 2.265 0.000 2.839 0.000 2.279 0.000
*
Pseudo-R-Squared 0.03 0.60 0.29
*
R2ML = 1 – exp(-G2/N), where G2 = -2 ln [L(Mα)/L(Mβ)]; Mα = restricted likelihood, Mβ = Unrestricted Likelihood,
and N=Number of observation (Maddala. 1983).

41
November 2011 Volume 42, Issue 3