Content uploaded by Sherrie R. Whaley
Author content
All content in this area was uploaded by Sherrie R. Whaley on Oct 11, 2017
Content may be subject to copyright.
Original article
Consumer perceptions of food-related risks
Mark Tucker,
1
* Sherrie R. Whaley
2
& Jeff S. Sharp
3
1 Agricultural Communication, Department of Human and Community Resource Development, The Ohio State University,
203 Agricultural Administration Building, 2120 Fyffe Road; Columbus, OH 43210-1067, USA
2 Grady College of Journalism and Mass Communication, University of Georgia, GA, USA
3 Rural Sociology, Department of Human and Community Resource Development, The Ohio State University, OH, USA
(Received 9 August 2003; Accepted in revised form 14 January 2005)
Summary Recent research indicates that the majority of respondents in nineteen of thirty-four
countries feel their food is less safe than 10 years ago. Concerns over food safety may result
in elevated levels of perceived risk, particularly when fuelled by intense media coverage.
The purpose of this study was to assess OhioansÕperceptions of various food safety risks
and to identify factors influencing risk judgments. Mail survey data are reported for 4014
respondents with a total response rate of 56%. Findings reveal moderate perceived levels of
risk for the food safety items assessed. Pesticide residues in food and contamination of
drinking water generated the highest levels of perceived risk, while mad cow disease and
genetically modified foods generated the lowest levels of perceived risk. Regression results
indicate that attitude toward biotechnology was the strongest predictor of perceived risk,
followed by perceptions of media system dependency. Findings from this research can
assist food safety specialists in developing more effective education and risk communi-
cation programmes for target audiences.
Keywords Biotechnology, consumer education, food safety, mass media, risk perceptions.
Introduction
Food safety is a growing concern in most of the
world. A survey conducted by the international
market research firm Ipsos-Reid (2000) found that
the majority of respondents in nineteen of thirty-
four countries feel their food is less safe than it was
10 years ago. With recent threats of food contam-
ination, consumers are increasingly interested in
the origin and content of their food (Butler, 2002).
Major interest groups involved with the food
system include consumers, industry, scientists, the
media and government officials (Sobal & Maurer,
1995).
Serious gaps in communication and under-
standing exist among these groups, according to
Thomson & Maretzki (1998), who argue that these
gaps contribute to consumersÕlimited understand-
ing of Ôfood and where it comes fromÕand to the
public’s growing concern and declining trust in
their food supply. Consumer concerns have
recently focused on such issues as genetically
modified foods (Gillam, 2002; Pew Initiative on
Food and Biotechnology, 2002); bacterial and
pesticide contamination (Sachs et al., 1987; Bryan,
1989; McIntosh et al., 1994; Chipman et al., 1996;
Food Marketing Institute, 2002; Whaley &
Doerfert, 2003); use of growth hormones in
livestock (Powell & Leiss, 1997); Ômad cowÕdisease
(Poulsen, 1996; Ten Eyck, 2000); and bio-terror-
ism (Klapthor, 2001).
Stenholm & Waggoner (1992) argue that con-
sumers will be the ultimate judges of food
products and processes and will determine their
merits, successes and failures. Therefore, it is
*Correspondent: Fax: 614 292 7007;
e-mail: tucker.9@osu.edu
International Journal of Food Science and Technology 2006, 41, 135–146 135
doi:10.1111/j.1365-2621.2005.01010.x
Ó2005 Institute of Food Science and Technology Trust Fund
important to gain a more complete understanding
of how the public forms risk judgments about
food safety. The purpose of this study was to
assess OhioansÕperceptions of food safety risks
and the factors influencing these perceptions.
Assessing consumersÕperceptions of risk related
to food safety serves multiple purposes. Empirical
measures of perceived public risks are needed by
policymakers and legislators to enact programmes
and policies to manage food safety risks and
maintain public confidence in the food supply
(Rosati & Saba, 2004). Information derived from
this research may also be used by food safety
specialists to develop effective education and risk
communication programmes for target audiences
(Blaine et al., 2002).
Theoretical approach and hypotheses
A theoretical approach was formulated from
selected components of risk-analysis theory and
media dependency theory to guide the investi-
gation. Risk-analysis theory acknowledges that
laypersons differ from experts in the manner by
which they form risk judgments. For instance,
experts tend to form risk judgments by considering
the potential severity or harm that might result
from an event as well as the probability that an
event may or may not occur (Slovic, 1987; Frewer,
1999). Laypersons, in contrast, often de-emphasize
the concept of probability when forming risk
judgments and rely more heavily on qualitative
aspects of the risk event, such as whether it can
be controlled, avoided, or easily understood
(Covello & Johnson, 1987; Pidgeon & Beattie,
1998; Bennett, 1999).
Individuals who perceive they are in control of a
particular hazard are less likely to attribute risk to
that situation. Such individuals are more likely to
focus on positive aspects of their behaviour that
can help reduce risk, rather than focus on their
own deficiencies or lack of control (Strecher et al.,
1986). In the food-safety context, individuals may
exercise control by making more careful and
informed food selection and preparation decisions.
However, those who perceive less control may be
more unlikely to undertake such proactive
measures. Therefore, it is expected that lower
levels of perceived control would be associated
with higher levels of perceived food safety risk.
The archetype example of lack of control is the
situation in which a person actually experiences
some type of food-borne illness. Individuals who
have suffered the negative consequences of food-
borne illness often become more vigilant about the
origin and proper preparation of foods they eat
and more sensitive to possible risks. Tolerance of
risk has also been shown to be lower for hazards
due to human error or carelessness rather than
those that occur from natural causes (Pidgeon &
Beattie, 1998). Because many food poisoning
incidents occur as a result of improper food
handling, storage or preparation, it is expected
that an individual’s experience with food poison-
ing would be associated with higher levels of
perceived food safety risk.
Risk-analysis literature asserts that perceptions
of risk are unevenly distributed across societies
(Short, 1984; McIntosh et al., 1994; Dosman
et al., 2001). Individuals who have greater access
to financial and educational resources tend to
express higher tolerances for risk, while those with
lower incomes and less education generally indi-
cate lower risk tolerances for most hazard situa-
tions (Tomazic et al., 2002). Individuals with
higher disposable incomes would be expected to
have greater opportunities to learn about and
follow recommended food safety practices, while
those with more education would be expected to
have greater awareness of food safety issues and
protective practices (Napier et al., 2004). There-
fore, it is expected that lower household incomes
and lower levels of education would be associated
with higher levels of perceived food safety risk.
The uneven distribution of risks in society is
also reflected in differential levels of perceived
risks expressed by various racial groups. Racial
minority groups have been shown to express lower
levels of trust in medical authorities as well as
lower tolerances of health and environmental risks
(Corbie-Smith et al., 2002; Johnson, 2002). In the
current analysis, it is expected that racial minority
status would be associated with higher levels of
perceived food safety risk.
Risk-analysis research has also shown that risk
situations that are easily understood by individu-
als are less likely to generate fear or dread than
risk situations that are not easily understood
(Pidgeon & Beattie, 1998; Frewer, 1999). Un-
familiar hazards are particularly worrying to
Factors influencing perceived food risk M. Tucker et al.136
International Journal of Food Science and Technology 2006, 41, 135–146 Ó2005 Institute of Food Science and Technology Trust Fund
individuals because they lack the knowledge
necessary to anticipate or avoid negative impacts,
some of which may be severe. In the current
analysis, it would be expected that individuals who
do the majority of cooking in their households
would be more knowledgeable about food safety
issues than those who do not because of their
greater experience with and repeated access to
food information. It is therefore expected that lack
of cooking responsibility would be associated with
higher levels of perceived food safety risk.
The theory also suggests that individuals are
more likely to seek food safety information and
increase their knowledge of food safety issues
when there are children in the household (Bennett,
1999; Dosman et al., 2001). In a Scottish study,
Reilly (1999) found that having children was one
of the most important incentives in changing food
consumption patterns. Because of the extra
demands and care required in preparing food for
children, it is expected that the presence of
children in the household would be associated
with higher levels of perceived food safety risk.
Thorough understanding of a particular risk
implies knowledge not only of the hazard itself,
but also of techniques or technologies that offer
avenues to minimize risk. In the case of food
safety, a great deal of media coverage has
surrounded biotechnology and its potential to
dramatically enhance food quality and availability
(Blaine et al., 2002). Despite its promise, biotech-
nology does not enjoy universal public support,
due partially to concerns about unknown health
and environmental dangers as well as ethical
dilemmas arising from the genetic modification
of plants and animals (Seidler et al., 1998; Napier
et al., 2004). Given the widespread public debate
about the risks and benefits of food biotechnology
today, it would be expected that individuals who
support biotechnology would express greater opti-
mism about food safety than those who do not
support it. Therefore, it is expected that more
negative attitudes toward biotechnology would be
associated with higher levels of perceived food
safety risk.
In addition to being influenced by formal
education and recent information seeking, risk
perceptions are also influenced by the totality of
experiences an individual has throughout his or
her lifetime. Individuals who have had significant
direct experience with or exposure to agricultural
production have been shown to express more
confidence in the food supply than their nonfarm
counterparts (Tomazic et al., 2002). While <2%
of the American population is actively engaged in
farming, a considerably higher proportion was
raised on or in close proximity to a farm. Such
individuals would be expected to perceive fewer
threats to food safety as a result of their greater
familiarity with food production, regardless of
their current residence (Napier et al., 2004).
Therefore, it is expected that individuals who were
raised in nonfarming areas with little opportunity
for exposure to agriculture would express higher
levels of perceived food safety risk.
The theoretical perspective used in this paper
also emphasizes the important role of mass media
in influencing individualsÕrisk judgments. Accord-
ing to Nelkin (1987), reliance on the media helps
to define the public’s sense of reality and its
perceptions of risks or benefits. Mass media are
consistently ranked by the public as a primary
source of food safety information (McIntosh
et al., 1994; Fisher & Chen, 1996; Borra et al.,
1998; Pisano & Woods, 2002).
To capture the role of media in influencing
public perception, this research uses elements of
media dependency theory to help identify predic-
tors of perceived food safety risk. A major
assertion of media dependency theory is that
individuals form strong attachments to mass
media to help satisfy their various needs for
information and entertainment (Littlejohn, 1996).
Siegrist (1999) argues that individuals must
increasingly rely on expert sources and institutions
for information about the potential risks and
benefits associated with complex new technologies,
such as gene technology, with which they have no
direct personal experience. Dependency on news
and information is also heightened during periods
of intense conflict or change. During such periods,
media respond with additional stories and cover-
age of interest to their audiences. Previous
research has found positive relationships between
media exposure and perception of health and
other general risk situations (Coleman, 1993;
Loges, 1994).
In the case of food safety issues, it is
important to note that media coverage often
focuses primarily on possible problems with the
Factors influencing perceived food risk M. Tucker et al. 137
Ó2005 Institute of Food Science and Technology Trust Fund International Journal of Food Science and Technology 2006, 41, 135–146
food supply as well as controversies related to
human health or nutrition. Intense media cover-
age of food safety issues can escalate into food
scares, particularly when food or food processes
are asserted to contain new or unexpected health
risks (Whelan & Stare, 1992; Craven & Johnson,
1999). Food safety coverage tends to cluster
around crisis situations, according to Ten Eyck
(2000). Given these findings, it is expected that
increased dependency on mass media systems
would be associated with higher levels of
perceived food safety risk (Whaley & Tucker,
2004).
To summarize, the application of risk analysis
and media dependency theories in the current
analysis has resulted in the following hypotheses:
H1: Individuals who perceive having less control
over the safety of their food are more likely
to perceive higher levels of food safety risk
than individuals who perceive having more
control.
H2: Individuals who have experienced food poi-
soning are more likely to perceive higher
levels of food safety risk than individuals
who have not experienced food poisoning.
H3: Individuals with lower household incomes
are more likely to perceive higher levels of
food safety risk than individuals with higher
household incomes.
H4: Individuals with less formal education are
more likely to perceive higher levels of food
safety risk than individuals with more edu-
cation.
H5: Individuals from racial minority groups are
more likely to perceive higher levels of food
safety risk than non-minority individuals.
H6: Individuals not responsible for cooking are
more likely to perceive higher levels of food
safety risk than individuals who are respon-
sible for cooking.
H7: Individuals with children in their households
are more likely to perceive higher levels of
food safety risk than individuals without
children in their households.
H8: Individuals who feel that biotechnology has
had a negative impact on the food supply are
more likely to perceive higher levels of food
safety risk than individuals who do not feel
that biotechnology has had a negative
impact.
H9: Individuals who were raised in a city, town,
or suburb removed from agricultural pro-
duction are more likely to perceive higher
levels of food safety risk than individuals
who were raised on a farm or in the country.
H10: Individuals who express greater levels of
media system dependency are more likely to
perceive higher levels of food safety risk than
individuals who express lower levels of
media system dependency.
Methods
Sampling and data collection
The sampling group of 7976 Ohioans was
provided by a private vendor. The sample was
stratified according to the state’s five Extension
districts and metropolitan core county status.
This manner of sampling was undertaken to
permit comparisons among respondents accord-
ing to particular regions and selected demo-
graphic characteristics that might otherwise be
underrepresented if random sampling were used.
For the current analysis, data were weighted to
permit state-level reporting.
With a population of more than 11.3 million,
Ohio is the seventh most populated state in the
US. The state is geographically and demograph-
ically diverse, with large farming areas and small
towns interspersed among the major urban centres
of Cleveland, Columbus and Cincinnati. Adding
to the state’s value as a source of attitudinal
research is its national reputation as a political and
consumer product test market. These characteris-
tics make Ohio an optimum study area for public
perceptions of food safety.
Mail survey techniques were used to collect data
for this study. Using elements of Dillman’s (2000)
tailored design method, the researchers made up
to five contacts with respondents. A prenotifica-
tion letter was mailed during the summer of 2002
explaining the purpose of the study and encour-
aging participation. The follow-up mailing consis-
ted of a cover letter, questionnaire, and a self-
addressed business-reply envelope. Two $1 bills
were included in this package as an incentive to
increase response. This mailing was followed over
several weeks with a reminder postcard, a second
survey questionnaire package, and a second
Factors influencing perceived food risk M. Tucker et al.138
International Journal of Food Science and Technology 2006, 41, 135–146 Ó2005 Institute of Food Science and Technology Trust Fund
reminder postcard. A total response rate of 56%
was achieved.
Measurement of study variables
The dependent variable in this study, perceived
food safety risk, was created by summing seven
items designed to tap respondentsÕperceived risk
toward various food safety issues. The items
were as follows: bacterial contamination, con-
tamination of drinking water, genetically modi-
fied foods, mad cow disease, pesticide residues in
food, growth hormones in meat or milk, and
bio-terrorist attacks on the food supply. Possible
responses ranged from 1, indicating no risk, to
7, indicating serious risk. Item analysis was
conducted to assess the reliability of the seven
items. The resulting alpha coefficient was 0.91,
indicating a high degree of internal consistency
among the items and justifying their use in a
composite measure (Cronbach, 1951; Mueller,
1986).
Ten independent variables were identified from
the theoretical perspective used to guide the study.
The variables were operationalized as follows:
Perceived control was measured by asking
respondents to indicate their level of agreement
with the statement, ÔI have little or no control over
the safety of my food.ÕResponses ranged from 1,
indicating strongly disagree, to 5, indicating
strongly agree. Midrange responses of 3 indicated
the respondent was undecided.
Incidence of food poisoning was measured by
asking respondents if anyone in their household
had ever experienced food poisoning. Responses
of yes were assigned a value of 1, and responses of
no were assigned a value of zero.
Household income was measured by asking
respondents to indicate their gross household
income for 2001. Response categories ranged from
1, indicating <$9999 of gross household income,
to 7, indicating $100 000 or more.
Education was measured by asking respondents
to indicate the number of years of education
completed at the time of the study.
Race was measured by asking respondents to
select one of six categories that best described
them. A value of 1 was assigned to respondents
describing themselves as African American, Asian,
Hispanic/Latino, Native American/American
Indian, or other. A zero was assigned to white
respondents.
Responsibility for cooking was measured by
asking respondents to indicate who was generally
responsible for cooking in their household.
Respondents indicating they were responsible for
cooking or indicating they shared the responsibil-
ity with a spouse or partner were assigned a value
of 1. Respondents who indicated that a spouse or
partner was generally responsible for cooking or
who selected the ÔotherÕresponse category were
assigned a value of zero.
Children in the household was measured by
asking respondents to indicate the number of
persons in their household under 5 years of age
and between 5 and 18. Responses for these two age
categories were combined and recoded so that the
presence of any persons in the household 18 years
of age or younger resulted in an assigned value of
1. A zero was assigned if no persons 18 years of
age or younger were reported in the household.
Perceived role of biotechnology was measured
by asking respondents to indicate their level of
agreement with the statement, ÔBiotechnology is
having a negative impact on the safety of our food
supply.ÕResponses ranged from 1, indicating
strongly disagree, to 5, indicating strongly agree.
Midrange responses of 3 indicated the respondent
was undecided.
Childhood residence in farming area was
measured by asking respondents to indicate in
what kind of place they spent most of their
childhood. The five possible responses were city,
suburb, small town, countryside, and farm.
Respondents indicating city, suburb, or small
town were assigned a value of 1, while those
indicating countryside or farm were assigned a
value of zero.
Media dependency was measured by asking
respondents to rate six mass media channels
according to their helpfulness in providing infor-
mation useful for taking care of the family and
running the household. The six media channels are
as follows: television news, television talk shows,
radio, newspapers, world wide web, and maga-
zines. Possible responses ranged from zero, indi-
cating not helpful, to 5, indicating very helpful.
Responses to the six items were summed to form a
composite measure of media dependency. Item
analysis was conducted to assess reliability of the
Factors influencing perceived food risk M. Tucker et al. 139
Ó2005 Institute of Food Science and Technology Trust Fund International Journal of Food Science and Technology 2006, 41, 135–146
composite variable, resulting in an acceptable
alpha coefficient of 0.80.
Statistical analysis
Data were analysed using the Statistical Package
for the Social Sciences. Descriptive statistics,
including frequencies, percentages and means,
were used to summarize responses and provide a
general summary of respondentsÕdemographic
characteristics. Multiple regression analysis was
used to test the theoretical model developed to
guide the study. Missing data were attributed the
variable mean as recommended by Donner (1982).
Results
Table 1 provides general information on
respondentsÕdemographic and attitudinal char-
acteristics relative to the study. Respondents
were generally middle-aged and reported com-
pleting an average of 2 years of education
beyond high school. Approximately 53% of the
respondents were females. A large majority
(89%) of the respondents were white. Gross
household income was widely distributed among
respondents. The modal income category was
$50 000–75 000. Approximately 12% of the
respondents reported household income of
$100 000 or greater.
Relative to respondentsÕresidence during child-
hood, nearly one-third (31%) indicated spending
most of their childhood in a city setting. Less than
one-fourth (23%) of the respondents reported
spending most of their childhood in the country-
side or on a farm.
Nearly 40% of the respondents reported chil-
dren in the household 18 years or less. Approxi-
mately one-fourth of these respondents reported
children in the household under the age of five.
Also reported in Table 1 are descriptive data for
respondentsÕlevel of responsibility for food pre-
paration in their household. Just over half (53%)
of the respondents reported having the primary
responsibility for food preparation. Nearly half
(48%) of the respondents reported that someone
in their household had experienced food poi-
soning.
Table 2 provides descriptive data on respond-
entsÕperceptions of food safety risk. As shown,
mean responses ranged from 4.56 to 5.33 on the
seven-point scale, indicating that respondents
perceived average to moderate levels of risk for
all of the items assessed. Pesticide residues in food
and contamination of drinking water generated
the highest levels of perceived risk, while mad cow
disease and genetically modified foods generated
the lowest levels. More than 30% of the respond-
ents perceived all of the risks except genetically
modified foods as serious.
Inspection of the standard deviations in Table 2
reveals relatively higher disagreement among
respondents regarding risks posed by bio-terrorist
Table 1 Demographic and other selected characteristics of
study respondents (n¼4014)
Age (year)
Median 50
Mean 51.2
SD 16.4
Education (year)
Mean education completed 14.0
SD 2.7
Gender (%)
Female 52.6
Male 47.4
Race/ethnicity (%)
White 89.4
African American 6.3
Asian 1.3
Hispanic/Latino 1.0
Native American/American Indian 0.9
Other 1.1
Gross household income (%)
<$9999 4.1
$10 000–19 999 12.1
$20 000–34 999 19.3
$35 000–49 999 17.4
$50 000–74 999 23.5
$75 000–99 999 11.9
$100 000 or more 11.8
Residence during childhood (%)
City 31.4
Suburb 21.3
Small town 22.8
Countryside 10.9
Farm 12.4
Children in household (%)
Households with children under age 5 10.9
Households with children 5–18 28.1
Responsible for food preparation (%)
You 52.5
Spouse/partner 21.4
Shared with spouse/partner 22.5
Other 2.7
Experienced food poisoning in household (%) 48.4
Factors influencing perceived food risk M. Tucker et al.140
International Journal of Food Science and Technology 2006, 41, 135–146 Ó2005 Institute of Food Science and Technology Trust Fund
attacks on the food supply and mad cow disease.
Table 3 provides data on respondentsÕperceived
helpfulness of various media in providing infor-
mation. Mean values ranged from 2.25 to 3.30 on
the six-point scale, indicating slight to moderate
levels of perceived helpfulness for the six media
channels assessed. Traditional media such as
newspapers and television news were assessed
most favourably. The relatively low standard
deviation for newspapers indicates relative agree-
ment among respondents as to the perceived
helpfulness of this medium.
Descriptive data for respondentsÕperceived level
of trust in various information sources are provi-
ded in Table 4. Mean values ranged from 3.30 to
3.91 on the five-point scale, indicating nearly
Table 2 Perceived risk to food safety posed by various issues, presented in percentages (n¼4014)
Issue
Level of risk
MD Mean SD
None Some Moderate Serious
1 234567
Pesticide residues in food 0.9 6.4 9.4 11.0 18.0 20.6 32.6 1.0 5.33 1.62
Contamination of drinking water 1.9 7.0 9.5 9.4 18.5 17.7 35.2 0.8 5.31 1.71
Growth hormones in meat or milk 2.0 6.4 8.7 12.7 16.9 18.5 33.3 1.4 5.28 1.69
Bacterial contamination 1.3 7.2 11.1 10.8 19.2 18.2 31.2 1.0 5.21 1.68
Bio-terrorist attacks on food supply 4.2 9.9 8.9 11.0 12.1 14.2 38.8 1.0 5.17 1.93
Mad cow disease 7.3 14.1 11.5 12.0 12.2 10.6 31.0 1.3 4.66 2.07
Genetically modified foods 4.7 10.0 11.9 21.3 17.3 14.7 18.2 2.0 4.56 1.76
Table 3 Perceived helpfulness of various media in providing news and information, presented in percentages (n¼4014)
Information medium
Not
helpful
Slightly
helpful
Moderately
helpful
Very
helpful
MD Mean SD012345
Newspapers 4.2 6.2 11.9 26.7 31.8 17.4 3.30 1.31
Television news 6.2 8.0 12.0 23.8 27.5 21.3 1.2 3.24 1.45
Magazines 8.6 10.1 16.8 29.1 23.8 9.9 1.7 2.80 1.41
Radio 10.4 10.7 17.1 24.6 23.4 11.6 2.2 2.76 1.49
World wide web 18.5 8.6 13.2 22.5 20.6 12.1 4.4 2.57 1.67
Television talk shows 22.0 13.7 16.8 19.5 16.8 9.7 1.6 2.25 1.66
Table 4 Perceived trust in various sources of information about environmental and food safety issues, presented in
percentages (n¼4014)
Information source
Level of trust
MD Mean SD
None Low Moderate High
1234 5
Physician/other health professional 1.2 6.2 15.8 51.5 23.1 2.3 3.91 0.87
University scientist 2.3 7.1 23.9 45.8 18.6 2.3 3.73 0.93
Farmer or grower 1.4 5.5 27.5 47.4 15.6 2.5 3.72 0.85
US Department of Agriculture 2.7 9.6 25.2 43.7 16.6 2.2 3.63 0.97
Extension educator/agent 2.4 8.0 32.5 40.8 12.7 3.6 3.55 0.91
US Food and Drug Administration 4.6 12.1 24.7 39.0 17.7 2.0 3.54 1.06
Friends or family 2.6 15.0 29.8 33.9 16.1 2.7 3.47 1.02
US Environmental Protection Agency 5.7 13.3 25.4 37.5 16.0 2.1 3.46 1.09
Consumer advocacy group 5.2 15.6 31.4 34.8 10.0 2.9 3.30 1.03
Factors influencing perceived food risk M. Tucker et al. 141
Ó2005 Institute of Food Science and Technology Trust Fund International Journal of Food Science and Technology 2006, 41, 135–146
moderate levels of perceived trust for all of the
sources assessed. Expert sources such as physi-
cians, health professionals, and university scien-
tists were evaluated most favourably, while friends
and family, the US Environmental Protection
Agency, and consumer advocacy groups were
evaluated least favourably.
Data regarding respondentsÕperceived level of
control over food safety are provided in Table 5.
Approximately 40% of the respondents indicated
they had little or no control over the safety of their
food, while 18% indicated they were undecided
about the issue. Also presented in Table 5 are
respondentsÕattitudes toward biotechnology. Re-
sults indicate that well over half (58%) of the
respondents are undecided as to the effect that
biotechnology is having on the food supply.
Approximately 20% of the respondents agreed
with the statement that biotechnology is having a
negative impact on food safety.
Multivariate findings
Multiple regression analysis with blockwise vari-
able entry was used to evaluate the utility of the
theoretical perspective used to guide the study
(Pedhazur, 1982). With this variable-entry
method, all independent variables were entered
into the model in a single step.
Before conducting the regression analysis, a
correlation matrix of independent variables was
inspected to detect any excessive intercorrelations
that could bias the regression findings. Based on
this procedure, gross household income and edu-
cation were observed to be moderately intercorre-
lated. Gross household income was therefore
eliminated from the model to avoid the threat of
multicollinearity. Variance in Perceived Food
Safety Risk was regressed against the remaining
nine independent variables. Findings are presented
below in standardized regression form. Independ-
ent variables designated with an asterisk are
significant at the 0.05 level.
Y¼0:263x
1þ0:180x
20:143x
3þ0:092x
4
0:062x
5þ0:052x
6þ0:027x70:021x8
þ0:018x9
Model F-test: 86.198 (P< 0.05).
Adjusted R
2
¼0.160.
Where Y, perceived food safety risk (scale); x
1
,
perceived role of biotechnology; x
2
, media
dependency (scale); x
3
, education; x
4
, race; x
5
,
children in the household; x
6
, incidence of food
poisoning; x
7
, perceived control; x
8
, responsibil-
ity for cooking; x
9
, childhood residence in
farming area.
Multivariate findings indicate that perceived
effect of biotechnology on the food supply was
the best predictor, followed by respondentsÕrepor-
ted media dependency, level of education, race,
presence of children in the household, and self-
reported incidence of food poisoning. The six
variables explained approximately 16% of the
variance in the dependent variable, indicating that
the model was of limited utility in explaining
respondentsÕperceived food safety risk.
Respondents expressing negative evaluations
of biotechnology were more likely to perceive
higher levels of food safety risk, as were those
expressing greater levels of media dependency.
Respondents with higher levels of education
tended to perceive lower levels of perceived risk,
as did white respondents and those reporting at
least one person in the household 18 years of
age or younger. Respondents who reported food
poisoning of someone in their household tended
to report higher levels of food safety risk.
Table 5 Attitudes toward perceived control of food safety and benefits of biotechnology, presented in percentages (n¼4014)
Statement
Strongly
disagree Disagree Undecided
Strongly
agree Agree
MD Mean SD123 45
I have little or no control
over the safety of my food
8.4 32.4 17.8 29.6 10.5 1.4 3.01 1.81
Biotechnology is having a
negative impact over the safety of my food
3.8 15.9 57.7 14.5 5.9 2.2 3.03 0.84
Factors influencing perceived food risk M. Tucker et al.142
International Journal of Food Science and Technology 2006, 41, 135–146 Ó2005 Institute of Food Science and Technology Trust Fund
Discussion
Descriptive and multivariate findings provide a
general profile of respondentsÕperceived level of
food safety risk, as well as some of the variables
that influence these perceptions. Respondents
generally expressed moderate levels of perceived
risk for all of the food safety issues assessed in this
study. The two items generating the highest levels
of perceived risk – pesticide residues in food and
contamination of drinking water – are often cited
as negative consequences of agricultural produc-
tion, which is a major industry in Ohio. While
issues such as mad cow disease and genetically
modified organisms have received considerable
media attention in recent months, both received
the lowest risk ratings among the seven items
assessed. These findings differ from those of
Rosati & Saba (2004), who found that pesticide
residues, bacterial contamination, and mad cow
disease [bovine spongiform encephalitis (BSE)]
generated the highest levels of perceived risk
among food hazards they assessed. However, their
research used Italian consumers as the population,
which might well account for differential risk
perceptions because of the European experience
with BSE.
Despite some variability in perceived risk
among the seven items comprising the dependent
variable, it should be noted that respondents
expressed relatively similar perceptions of risk
for each of the items assessed. This finding is
somewhat unexpected given the fact that the items
vary greatly in their likelihood of occurrence and
in the steps that can be taken by consumers to
mitigate risk. For instance, although bacterial
contamination can occur in many different types
of foods, the danger can be significantly reduced
or eliminated by observing proper food prepar-
ation techniques. Conversely, consumers have
little personal control in avoiding the potentially
deadly consequences of mad cow disease or its
human variant, Creutzfeldt-Jakob disease,
although neither has resulted in human death or
illness in the United States to date.
The apparent lack of discrimination among
respondents for very different food risk situations
suggests some possible implications for profes-
sional communicators and food safety specialists.
Among these is the possibility that respondents are
not aware of the level of control they can exercise
to reduce certain food safety risks. Therefore, in
addition to preparing news stories and reports that
publicize instances and dangers of food contam-
ination, communicators should also continue to
stress the steps that individuals can take to reduce
the risk of contamination and food-borne illness.
In the cases of mad cow disease and genetically
modified organisms, communicators should stress
that there is no documented incidence of illness
from either of these sources in the United States.
While these practices are being followed in many
cases, findings from this study suggest that new or
additional information strategies should be ex-
plored to reach consumers more effectively.
One such strategy is the formation of more
formal working relationships between communi-
cators and food safety specialists to prepare
newsworthy mass media stories and publications
that will attract readership without creating undue
public concerns. Food safety specialists can play a
key role in this process by helping the media to
provide balance and perspective in their coverage.
By reviewing news stories and reports prior to
their release for public dissemination, food safety
specialists can help make food-related information
more understandable and relevant to both media
and consumers. The pooling of communications
and food safety expertise would likely result in the
development of more accurate and socially
responsible editorial products that have a greater
chance of being published by mass media.
Regression findings indicate that the theoretical
model developed in the study was of limited value.
Six of the nine hypothesized variables accounted
for approximately 16% of the variance in per-
ceived food safety risk. As expected, respondents
expressing higher levels of food safety risk tended
to view biotechnology as having a negative impact
on the food supply and expressed higher levels of
dependence on mass media channels. Also consis-
tent with theory was the finding that higher levels
of perceived risk were expressed by less-educated
respondents, racial minorities, and those reporting
an incidence of food poisoning in their household.
The finding that attitude toward biotechnology
was the best predictor of food safety risk is
particularly interesting because over half of the
respondents indicated they were undecided as to
whether biotechnology was having an overall
Factors influencing perceived food risk M. Tucker et al. 143
Ó2005 Institute of Food Science and Technology Trust Fund International Journal of Food Science and Technology 2006, 41, 135–146
negative or positive impact on the food supply.
Food safety specialists and communicators can
play a critical role in helping educate consumers
about biotechnology during a time when a major-
ity are forming opinions as to its risks and
benefits. Given the large proportion of consumers
who remain undecided on the merits of biotech-
nology, it is important that science-based infor-
mation be made available via commonly used
mass media channels and presented in an under-
standable manner. According to Irani et al.
(2002), such information should focus not only
on the science associated with biotechnology, but
also on such topics as the accountability and
regulation of the biotechnology industry, which
may also greatly influence consumer acceptance of
food biotechnology.
The finding that food safety risks are elevated
as media dependency increases is also deserving
of additional consideration. It is important to
note that food safety programmes are often
justified on the basis that they educate and
empower audiences by providing useful informa-
tion to help avert risks. Under these circum-
stances, it would be expected that increased
reliance on mass media for food safety informa-
tion would result in lower levels of perceived
risk. The finding that increased reliance on food
safety information tends to result in higher levels
of perceived risk raises some question as to the
effectiveness of current communication pro-
grammes. If the positive relationship between
perceived risk and media system dependency is
found to persist over time, then adjustments
would seem to be warranted in how food safety
information is delivered to consumers. Such a
finding would particularly support the recom-
mendation provided earlier to forge more func-
tional working relationships between food safety
specialists and professional communicators to
help improve mass media coverage of food
safety issues.
The finding that respondents with children in
the household perceived lower levels of risk than
respondents with no children in the household was
unexpected. One possible explanation for the
finding is that individuals with children in the
household have a greater stake in ensuring they
are knowledgeable about food safety issues. Such
individuals might engage in more information
seeking about food safety risks than those without
children in the household and gain greater aware-
ness of how to avert risks. Future research should
investigate this possibility using different study
populations.
Three variables (perceived control, responsibil-
ity for cooking, and childhood residence in farm-
ing area) were statistically insignificant as
predictors of perceived food safety risk. One
possible reason for the failure of perceived control
to enter the model is the manner in which the
dependent variable was operationalized. In this
study, the dependent variable was a composite
measure of seven items representing quite different
food safety issues that vary greatly according to
their level of control. It is possible that perceived
control would have entered the model as statisti-
cally significant if the dependent variable were
composed only of items that are more amenable to
individual control, such as bacterial contamin-
ation. The failure of responsibility for cooking to
enter the model may also be due to the nature of
the dependent variable as operationalized in this
study. Issues such as mad cow disease and bio-
terrorist attacks on the food supply have been
widely publicized by the general mass media. The
widespread media coverage and dissemination of
such information would likely lead to increasing
awareness of food safety issues by all consumers,
not just those actively seeking food safety infor-
mation. The increase in awareness of such issues
among all consumers might thus explain the
failure of cooking responsibility to enter the
model. Finally, the failure of childhood residence
to enter the model was unexpected, as it was
hypothesized that individuals with less exposure to
agriculture would express higher levels of per-
ceived food safety risk. Because childhood resi-
dence is only an indirect measure of agricultural
knowledge or experience, it is possible that the
statistical insignificance of the variable can be
traced to a measurement problem. However, if
alternative measures of agricultural knowledge fail
to enter models in future research, then it is
possible that perceptual differences between those
having and not having exposure to farm produc-
tion may not exist as hypothesized. If confirmed
through further empirical research, the finding
would necessitate adjustments in the theoretical
perspective used in the current study.
Factors influencing perceived food risk M. Tucker et al.144
International Journal of Food Science and Technology 2006, 41, 135–146 Ó2005 Institute of Food Science and Technology Trust Fund
Conclusions
Despite some limitations in the theoretical model
developed here, findings from this research
suggest that the fusion of risk-analysis and
media dependency theories may hold promise
in future investigations. Risk-analysis theory was
successful in its assertion that respondentsÕ
perceptions of biotechnology, various demogra-
phic characteristics, and experience with food
safety problems are positively associated with
perceived risk. Media dependency theory provi-
ded additional explanatory value to the model
by accounting for the role of mass media in
helping shape perceptions of risk.
A challenge for future research is to identify and
test other variables that might account for addi-
tional unexplained variance in consumer percep-
tions of food safety risk. As discussed earlier, it is
possible that the current model could be improved
by focusing the dependent variable on a more
specific food safety risk, rather than the general
measure used here. It is also possible that the
model might explain additional variance if the
research focused specifically on consumers who
are actively pursuing information about a parti-
cular food safety issue. Future research will help
answer these questions and provide additional
insights.
References
Bennett, P. (1999). Understanding responses to risk: some
basic findings. In: Risk Communication and Public Health
(edited by P. Bennett & K. Calman). Pp. 3–19. New York:
Oxford University Press.
Blaine, K., Kamaldeen, S. & Powell, D. (2002). Public
perceptions of biotechnology. Journal of Food Science,67,
3200–3208.
Borra, S.T., Earl, R. & Hogan, E.H. (1998). Paucity of
nutrition and food safety Ônews you can useÕreveals
opportunity for dietetics practitioners. Journal of the
American Dietetic Association,98, 190–193.
Bryan, F.L. (1989). Risks associated with vehicles of
foodborne pathogens and toxins. Journal of Food Protec-
tion,51, 498–508.
Butler, L.M. (2002). Rural–urban interdependency and the
future of agriculture. Paper presented at the Agricultural
Outlook Forum 2002, Washington, D.C.
Chipman, H., Kendall, P., Slater, M. & Auld, G. (1996).
Audience responses to a risk communication message in
four media formats. Journal of Nutrition Education,28,
133–139.
Coleman, C. (1993). The influence of mass media and
interpersonal communication on societal and personal
risk judgments. Communication Research,20, 611–628.
Corbie-Smith, G., Thomas, S.B. & St. George, D.M.M.
(2002). Distrust, race, and research. Archives of Internal
Medicine,162, 2458–2463.
Covello, V.T. & Johnson, B. B. (1987). The social and
cultural construction of risk: issues, methods, and case
studies. In: The Social and Cultural Construction of Risk
(edited by B. B. Johson & V. T. Covello). Pp. vii–xiii.
Dordrecht, Holland: D. Reidel Publishing Company.
Craven, B. & Johnson, C. (1999). Politics, policy, poisoning
and food scares. In: Fearing Food: Risk, Health and
Environment (edited by J. Morris & R. Bate). Pp. 141–
169. Oxford: Butterworth Heinemann.
Cronbach, L.S. (1951). Coefficient alpha and the internal
structure of tests. Psychometrika,16, 297–334.
Dillman, D.A. (2000). Mail and Internet Surveys: the
Tailored Design Method. New York: John Wiley & Sons.
Donner, A. (1982). The relative effectiveness of procedures
commonly used in multiple regression analysis for dealing
with missing values. The American Statistician,36, 378–
381.
Dosman, D.M., Adamowicz, W.L. & Hrudey, S.E. (2001).
Socioeconomic determinants of health- and food safety-
related risk perceptions. Risk Analysis,21, 307–317.
Fisher, A. & Chen, Y.C. (1996). Customer perceptions of
agency risk communication. Risk Analysis,16, 177–
184.
Food Marketing Institute. (2002). Trends in the United
States 2002: Consumer Attitudes and the Supermarket.
Pp. 41–45. Washington, D.C.: FMI.
Frewer, L.J. (1999). Public risk perceptions and risk
communication. In: Risk Communication and Public
Health (edited by P. Bennett & K. Calman ), Pp. 20–32.
New York: Oxford University Press.
Gillam, C. (2002). U.S. Consumers Challenge Spread of
Biotech Food. Available at: http://foodfirst.org/archive/
media/news/2002/consumerschallenge.html. Reuters News
Service.
Ipsos-Reid (2000). Food safety a growing concern in most of the
world. (Press release). [On-line]. Available: http://www.
ipsos-na.com/news/pressrelease.cfm?id¼1195#.
Irani, T., Sinclair, J. & O’Malley, M. (2002). The importance
of being accountable. Science Communication,23, 225–242.
Johnson, B.B. (2002). Gender and race in beliefs about
outdoor air pollution. Risk Analysis,22, 725–738.
Klapthor, J.N. (2001). Food security and food safety:
Today’s topics of news media. Food Technology,55, 20.
Littlejohn, S.W. (1996). Theories of Human Communication.
Pp. 324–354. Belmont, CA: Wadsworth Publishing Co.
Loges, W.E. (1994). Canaries in the coal mine: perceptions
of threat and media system dependency relations. Com-
munication Research,21, 5–23.
McIntosh, W.A., Acuff, G.R., Christensen, L.B. & Hale, D.
(1994). Public perceptions of food safety. The Social
Science Journal,31, 285–292.
Mueller, D.J. (1986). Measuring Social Attitudes. Pp. 8–19.
New York: Teachers College Press.
Factors influencing perceived food risk M. Tucker et al. 145
Ó2005 Institute of Food Science and Technology Trust Fund International Journal of Food Science and Technology 2006, 41, 135–146
Napier, T.L., Tucker, M., Henry, C. & Whaley, S.R. (2004).
Consumer attitudes toward GMOs: The Ohio experience.
Journal of Food Science,69, 69–76.
Nelkin, D. (1987). Selling Science: How the Press Covers
Science and Technology. New York: W.H. Freeman.
Pedhazur, E.J. (1982). Multiple Regression in Behavioral
Research. Forth Worth: Holt, Rinehart and Winston.
Pew Initiative on Food and Biotechnology (2002). Consum-
ers Evenly Divided Over Environmental Risks and Benefits
of Genetically Modified Food and Biotechnology. News
release [On-line] Available: pewagbiotech.org/newsroom/
releases/020402.php3
Pidgeon, N.F. & Beattie, J. (1998). The psychology of risk
and uncertainty. In: Handbook of Environmental Risk
Assessment and Management (edited by P. Calow).
Pp. 289–318. Oxford: Blackwell Science Ltd.
Pisano, C. & Woods, M.D. (2002). Communicating food
safety across cultures: Issues of trust and credibility
within diverse populations. Paper presented at the
international meeting of Agricultural Communicators in
Education, Savannah, GA.
Poulsen, J. (1996). When journalism loses its senses: on mad
cow disease and ritual sacrifice. Nordicom Review,17, 3–9.
Powell, D. & Leiss, W. (1997). Mad Cows and Mother’s
Milk: the Perils of Poor Risk Communication. Pp. 220–
221. Montreal: McGill-Queen’s University Press.
Reilly, J. (1999). ÔJust another food scare?ÕPublic under-
standing and the BSE crisis. In: Message Received:
Glasgow Media Group Research, 1993–1998 (edited by
G. Philo), Pp. 128–145. New York: Addison Wesley
Longman.
Rosati, S. & Saba, A. (2004). The perception of risks
associated with food-related hazards and the perceived
reliability of sources of information. International Journal
of Food Science and Technology,39, 491–500.
Sachs, C., Blair, D. & Richter, C. (1987). Consumer
pesticide concerns: a 1965 and 1984 comparison. Journal
of Consumer Affairs,21, 96–107.
Seidler, R.J., Watrud, L.S. & George, S.E. (1998). Assessing
risks to ecosystems and human health from genetically
modified organisms. In: Handbook of Environmental Risk
Assessment and Management (edited by P. Calow).
Pp. 110–146. Oxford: Blackwell Science Ltd.
Short, J. F. (1984). The social fabric at risk: toward the
social transformation of risk analysis. American Socio-
logical Review,49, 711–725.
Siegrist, M. (1999). A causal model explaining the percep-
tion and acceptance of gene technology. Journal of
Applied Social Psychology,29, 2093–2106.
Slovic, P. (1987). Perception of risk. Science,236, 280–285.
Sobal, J. & Maurer, D. (1995). Food, eating, and nutrition
as social problems. In: Eating Agendas: Food and
Nutrition as Social Problems (edited by D. Maurer &
J. Sobal ), Pp. 3–7. New York: Aldine De Gruyter.
Stenholm, C.W. & Waggoner, D.B. (1992). Public policy in
animal biotechnology in the 1990s: challenges and
opportunities. In: Animal Biotechnology: Opportunities
and Challenges (edited by J.F. MacDonald). Pp. 25–35.
Ithaca, NY: National Agricultural Biotechnology
Council.
Strecher, V. J., DeVellis, B. M., Becker, M. H. & Rosen-
stock, I. M. (1986). The role of self-efficacy in achieving
health behavior change. Health Education Quarterly,13,
73–91.
Ten Eyck, T. A. (2000). The marginalization of food safety
issues: an interpretative approach to mass media cover-
age. Journal of Applied Communications,84, 29–47.
Thomson, J.S. & Maretzki, A.N. (1998). Expanding the
Dialogue about the Food System: Using a Media Forum to
Enhance Communications Among Food System Stakehold-
ers. Project summary, Year 1, Nov. 1, 1997-Oct. 31, 1998.
Submitted to the Keystone 21 initiative of the W.K.
Kellogg Foundation. Unpublished manuscript.
Tomazic, T.J., Katz, B.M. & Harris, C.K. (2002). Is that
strawberry safe to eat? Consumer attitudes about food
safety. In: The Social Risks of Agriculture (edited by R.C.
Wimberly, C.K. Harris, J.J. Molnar & T.J. Tomazic).
Pp. 57–74. Westport, CT: Praeger Publishers.
Whaley, S.R. & Doerfert, D. L. (2003). Is Your Food Safe or
Scary? How U.S. News Magazines Communicated Food
Safety Issues, 1990–2000. Paper presented at the interna-
tional meeting of Agricultural Communicators in Educa-
tion, Kansas City, MO.
Whaley, S.R. & Tucker, M. (2004). The influence of
perceived food risk and source trust on media system
dependency. Journal of Applied Communications,88,9–
27.
Whelan, E.M. & Stare, F.J. (1992). Panic in the Pantry (2nd
edn). Pp. 1–16. Buffalo, N.Y.: Prometheus.
Factors influencing perceived food risk M. Tucker et al.146
International Journal of Food Science and Technology 2006, 41, 135–146 Ó2005 Institute of Food Science and Technology Trust Fund
