Specific Applications of Genetic

Public Concerns in the United
Kingdom about General and
Specific Applications of Genetic
Engineering: Risk, Benefit, and Ethics
Lynn J. Frewer
Chaya Howard
Richard Shepherd
Institute of Food Research, Reading, U.K.
The repertory grid method was used to determine what terminology respondents use to
distinguish between different applications of genetic engineering drawn from foodrelated, agricultural, and medical applications. Respondents were asked to react to
fifteen applications phrased in general terms, and results compared with a second study
where fifteen more specific applications were used as stimuli. Both sets of data were
submitted to generalized Procrustes analysis. Applications associated with animals or
human genetic material were described as causing ethical concern, being unnatural,
harmful, and dangerous. Those involving plants or microorganisms were described as
beneficial, progressive, and necessary. The results were validated in survey research,
which indicated that general applications ofgenetic engineering were perceived as either
positive or negative, whereas specific applications were more highly differentiated in
perceptual terms. The results imply that the public debate about genetic engineering must
take due account of the complexity of public concerns.
Public attitudes have been identified as one of the key determinants of the
development, application, and subsequent technological evolution of technology (Advisory Committee on Science and Technology 1990). Recent
research to understand public attitudes toward biotechnology and, more
specifically, genetic engineering has focused on attitudes toward the technol-
AUTHORS’ NOTE: This research was conducted as part of a project entitled &dquo;Detenninants of
Public Acceptance or Rejection of Genetic Engineering: Individual and Group Differences and
Effective Communication,&dquo; funded by the Biotechnology and Biological Sciences Research
Science, Technology, & Human Values, Vol. 22 No. 1, Wmter 1997 98-124
0 1997
Sage Publications Inc.
98
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
99
ogy overall, rather than its specific applications. Generally, this research has
relied on questions generated by the researchers attempting to determine
consumer attitudes toward the technologies. Typically, the opinion poll
methodology has been used (Zechendorf 1994) to ask simple questions
regarding acceptance of the technology. Although some evidence suggests
that currently the general public knows or understands little about biotechnology (Frewer, Shepherd, and Sparks 1994a; Hamstra 1991), public awareness is likely to increase through media and product exposure.
To understand what determines public concerns regarding the technology,
it would be advantageous not to make a priori assumptions about what the
public is likely to consider important. The aim of the present study was to
determine what underlying psychological constructs shape public perceptions of genetic engineering. Our methodology allowed us to investigate what
issues associated with the technology were most salient for the public. Data
were collected both when the technology was presented in a very general way
and when the respondents considered more specific applications of the
technology, whose tangible benefits were more obvious. Finally, the results
were validated in a larger survey, which again examined differences in the
way respondents thought about genetic engineering presented in a very
general way and about specific applications of the technology.
Past Research on Public
Attitudes Toward Biotechnology
DNA research is typically perceived as an unknown and moderately
dreaded hazard relative to other hazards (Slovic 1987). This general finding
has been confirmed for the food-related applications of biotechnology that
had been compared with other food-related hazards (Sparks and Shepherd
1994). Public concerns about the development and application of the technology are likely to reflect not only the scientific issues associated with the
technology (Harlander 1991) but also more fundamental ideas such as ethical
beliefs (Hoban, Woodrum, and Czaja 1992) and perceptions of unnaturalness
(Frewer, Howard, and Shepherd 1996).
Zechendorf (1994) has reviewed the results of the many surveys into
public opinion about biotechnology that have been conducted in Europe, the
United States, and Japan during the last ten years. Although there are many
Council in the United Kingdom. We thank Marguerite Fazey for assistance with the data
processing and Duncan Hedderly for assistance with the data analysis. We would also like to
thank an anonymous referee, whose suggestions have been incorporated in the discussion
section.
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
100
problems in comparing the results of surveys whose questions and methodologies differ, some overall trends were identified. For example, attitudes
appear to be related to the nature of the application. Medical applications are
more acceptable than food-related applications (particularly in Japan), and
applications involving plants or microorganisms are more acceptable than
those involving animals (although this trend is reduced in Southern European
countries).
Indeed, the importance of the link between application specificity and
attitude formation has been a common element in the literature. Of particular
relevance is the distinction between applications involving animals compared
with other applications (Hoban and Kendall 1992; Marlier 1992; Sparks,
Shepherd, and Frewer 1994).
There is some evidence that differences in attitude within the area of food
production depend on specific applications and their associated benefits,
whether products are presented as hypothetical developments (Hamstra
1993) or as realistic products (Frewer, Howard, and Shepherd 1996). Similarly, whether or not ethical concerns are raised depends on specific applications rather than the technology overall (Fox 1988; Harlander 1991;
Straughan 1992; Sparks and Shepherd 1996). It is becoming apparent that
public attitudes toward the technology overall are unlikely to define acceptance ; rather, the public will accept or reject applications of the technology
on a case-by-case basis.
It is important to compare perceptions associated with general applications with those associated with specific (and hence more tangible) applications. If public concern is defined by the nature of the application, then
perhaps the psychological constructs that underlie this concern are multidimensional, rather than defined by a single dimension such as perceived risk.
To predict public attitudes toward a specific application of genetic engineering, it is necessary to take due account of these relevant psychological
constructs and their importance for the public.
Ethical concerns play an important role in public reactions to genetic
engineering. Previous research, based on a modified version of Ajzen’s
(1988) theory of planned behavior, assessed beliefs about the likelihood of
possible outcomes of the technology (Sparks, Shepherd, and Frewer 1995).
The theory suggests that individuals act as a result of their behavioral
intentions, which are a function of their attitudes. Behavior is also influenced
by people’s perceptions of the social pressure put upon them and their beliefs
about the extent to which they can control actually performing the behavior
(Ajzen 1988). The results indicated that people’s attitudes are based not only
on instrumental considerations but also on ethical concerns.
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
101
The Social Construction
of Risk
People’s responses to general risk issues may be different from their
responses to specific risk issues. Dramatic or sensational risks tend to be
greatly overestimated, while less dramatic ones are underestimated (Dake
and Wildavsky 1991; Morgan et al. 1985). People may react differently
to abstract risks than to more concrete, specific hazards (Slovic 1992).
The way in which people characterize different hazards has been examined for specific domains (for example, automobile accidents; Slovic,
Macgregor, and Kraus 1987; and rail hazards, Kraus and Slovic 1988).
Risk perceptions are socially constructed (Jasanoff 1993), and disparity between lay and expert perceptions of risk for a variety of hazards has
been well documented (e.g., Flynn, Slovic, and Mertz 1993; Sjoberg and
Drotzz-Sjoberg 1994). This disparity can also be explained by the fact that
expert risk assessments do not take account of the social construction of
risk, while public reactions and public resistance to technological hazards
are likely to be shaped by factors other than scientific risk estimates
(Bauer 1995). Power-or the public perception of power-is likely to be
an important determinant of the debates about risks and of the public
perceptions of risks (Stallings 1990). The credibility of the risk regulators
is also likely to be an important factor in public risk acceptance (Slovic
1993). When exposure to a risk is perceived as involuntary, the risk is
regarded as more threatening than when an individual has a choice over
personal exposure (Sharlin 1989). What is apparent is that risk perceptions
cannot be viewed independently of the social context in which they are
embedded.
Hilgartner (1990) has commented on the cultural dominance of science
and scientists in the dissemination of simplified scientific results to the
public. He proposes that the one-way model of science communication
serves as a powerful tool in maintaining the social hierarchy of expertise.
Genuine scientific knowledge is regarded as the exclusive preserve of
scientists and is not accessible to either policymakers or the general
public. Scientific experts can then construct simplifications of science to
support their own goals. It is possible, however, that public attitudes
toward science in general, and genetic engineering in particular, are
psychologically far more complex than the simplified descriptions of
the relevant scientific processes suggest. It is necessary to understand the
complexity of these public concerns within the framework of science
communication.
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
102
The Repertory Grid and
Generalized Procrustes Analysis
Much of the risk perception literature uses the factor-analytic approach to
elicit the underlying psychological constructs that determine risk perceptions
(Fischhoff et al. 1978; Vlek and Stallen 1981). The principal criticism of this
approach is the use of experimenter-selected characteristics to describe the
psychological constructs involved. The attitudes that are measured are those
that the experimenter thinks are important at the outset of the experiment,
and they may have little meaning to the respondent, although he or she is still
required to make a response.
Heijs, Midden, and Drabbe (1993) have criticized much of the attitude
research because it does not provide insight into reasons underlying responses to questions. In addition, many surveys use abstract terminology
relating to biotechnology, which is likely to produce responses with only a
limited degree of validity because of the public’s lack of familiarity with this
terminology.
Much of the research that has examined public attitudes toward genetic
engineering has involved researcher-defined constructs (see Zechendorf
1994 for review). For example, such constructs have been an essential part
of the development of explanatory models of attitude and behavioral intention (Sparks, Shepherd, and Frewer 1995). Other research has assumed that
risk perceptions associated with genetic engineering will be determined by
the same underpinning psychological factors as those that define more global
risk domains (Frewer, Shepherd, and Sparks 1994b).
However, to understand why people adopt particular attitudes toward
genetic engineering and its applications, it is inappropriate to generalize from
global risk perception models, as it is possible that particular attitudinal
characteristics, such as ethical concerns, will contribute to public risk perceptions of genetic engineering but not to perceptions of other risks. Similarly, risk perceptions within some hazard domains may be defined by
psychological representations of risk that do not apply to genetic engineering.
Previous attempts to assess underlying determinants of attitudes that used
interviewing methods have been uninformative, partly because respondents
knew little about the technology (Frewer, Shepherd, and Sparks 1994a) and
partly because the research did not directly investigate psychological determinants of risk perceptions (Martin and Tait 1992). Open-ended interviewing
is unlikely to produce systematic comparative data that would allow for the
generation of predictive models because respondents are unlikely to focus
their concerns on the wide range of applications being developed.
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
103
A methodological solution to such problems is provided by the repertory
grid methodology (Kelly 1955) in conjunction with generalized Procrustes
analysis (GPA; Gower 1975). This methodology has been successfully used
in sensory evaluation studies (Gains and Thompson 1990; Guy, Piggott, and
Marie 1989; Raats and Shepherd 1991/1992) and in the development of
psychological models of the perceived risks associated with chemicals in
food (Raats and Shepherd 1996). The repertory grid method allows
respondents to describe their concerns using their own words and focuses
responses on the hazard domain under consideration without imposing
external, experimenter-determined risk characteristics on the data. The use
of GPA allows the differentiation of constructs about which respondents
agree, and the most important determinants of an attitude can be identified.
The repertory grid method involves randomly presenting respondents with
a set of three stimuli (in this case, potential applications of genetic engineering). For example, a respondent is presented with the following list: genetic
engineering of microorganisms (e.g., yeast) for food production purposes,
genetic engineering of plants for food production purposes, and genetic
engineering of animals for medical purposes. Respondents are then asked to
rank the three stimuli according so some criterion. In the case of research on
perceptions of risk, a typical question might be to rank the stimuli according
to the concern they cause. The respondent is then asked to explain why he or
she ranked the stimuli in this particular order. Interobserver reliability techniques are then applied to this semistructured interview data to determine the
relevant constructs (using the respondent’s own words wherever possible),
and a personalized questionnaire is developed for each respondent. Respondents are then asked to assess (typically, on continuous bipolar linear scales)
each of the initially presented stimuli against each of the constructs developed
in the repertory grid method. The entire data set is then subjected to GPA.
The application of GPA results in a graphical representation of a set of
stimuli (in this case, applications of genetic engineering) in several dimensions, each new dimension incorporating psychological constructs of importance from all the individuals interviewed. The most highly correlated constructs are deemed to be the most important dimensions in the resultant
model, accounting for a greater proportion of the resultant factor structure.
GPA is particularly useful when information is required about how individuals differ and to what extent they agree in their perceptions of the same
stimulus (Dijksterhuis and Gower 1991). The technique allows for differences in perception to be incorporated into the analysis (Arnold and Williams
1986).
The current research used the repertory grid method to examine the
psychological constructs underlying attitudes toward various applications of
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
104
genetic engineering drawn from food-related, agriculture-related, and mediareas. Two sets of investigations were conducted. The first used very
generalized abstract applications of genetic engineering-similar to those
used in previous research (Frewer and Shepherd 1995)-as stimuli for the
repertory grid analysis. The second used more specific stimuli that may have
a more realistic impact in terms of personal relevance of the technology and
tangible benefits (Frewer, Howard, and Shepherd 1996). We could then
compare responses to the two sets of stimuli and the underlying psychological
constructs shaping the perceptions of genetic engineering. Finally, to check
whether the results of the first two experiments were indeed representative
of the whole population, the results were validated by factor-analytic survey
research using larger respondent samples.
cal
Methods and Results
Study 1
The psychological determinants of attitudes toward general applications
of genetic engineering were elicited from twenty-five respondents, nineteen
females and six males (mean age = 47.5 years, SD = ±18.2 years), recruited
from the Reading area. Upon completion of the repertory grid interview,
respondents were paid £10 (approximately U.S.$15). Respondents were
drawn from a subject panel available in the Reading area and selected from
different ages and socioeconomic groups. Although it is not possible to quota
sample in such a small sample, the respondents were reasonably distributed
in terms of socioeconomic group and age.
In the repertory grid method, respondents were asked about fifteen possible general applications of genetic engineering in food production, medicine,
and agriculture (see Table 1). The repertory grid method was divided into two
phases. In the first phase, constructs describing concerns about the different
applications of genetic engineering were elicited from respondents. In the
second phase, respondents rated each of the applications on each construct
they had personally described as relevant.
All respondents were given a standardized definition of genetic engineering at the outset of the experiment.’ Each respondent was given a questionnaire with the applications of genetic engineering presented in groups of three
on separate pages, in fully randomized order. Each application was presented
twice within the questionnaire, to give ten different combinations of stimuli
in total (a similar method to that adapted from the repertory grid method by
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
105
Table 1.
Applications Used in the Generation of Constructs
about Genetic Engineering
Raats, Sparks, and Grugeon 1994). For each triad, respondents
&dquo;Which of these
applications
of
genetic engineering gives
were
asked,
you the most
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
106
why?&dquo; and &dquo;Which of these applications of genetic engineering
gives you the least concern, and why?&dquo; The respondents thus listed reasons,
attributes, or constructs that they used to rank the different applications of
genetic engineering. Responses were written down on the questionnaire by
the respondents themselves. From these data, a personalized questionnaire
was created for each respondent. The respondent scored each of the applications of genetic engineering for each construct on unstructured line scales
with personalized labeled end points derived from the elicitation.
The data were submitted to GPA (Gower 1975). The analysis used the
method proposed by TenBerge (1977). A GPA group average perceptual space
was obtained, illustrating the relative positions of the applications.
The repertory grid data were classified into one of forty-one classes by
two researchers. For example, if a respondent stated that he or she was
concerned with a particular application because it involved &dquo;fiddling about
with nature,&dquo; this was deemed to fall within the class of &dquo;tampering with
nature.&dquo; Classification initially used a predetermined category system, although not all categories were used. In addition, extra classes were added in
the course of the analysis. When disagreement occurred, the classification
concern, and
discussed until agreement was reached.
The resulting construct classifications are listed in alphabetical order in
Table 2. The number of constructs elicited from each respondent ranged from
4 to 23, with the mean number of constructs being 9.3. Figure 1 shows how
the different applications were perceived in relation to the constructs elicited
from the respondents. The first two axes of the GPA group average configuration together account for 75 percent of the variation in the data. Axis 3
accounted for 9 percent, and subsequent axes 5 percent or less. Interpretation
was therefore confined to the first two axes. Table 2 lists the number of
constructs that have high correlations with the first two axes. Only those
construct classes that occur more than five times are considered important.
The first axis explains 63 percent of the data variation. The terms used to
describe the genetic engineering applications on the side of axis 1 that refers
to more positive attitudes are &dquo;beneficial,&dquo; &dquo;advantageous,&dquo; &dquo;necessary,&dquo; and
&dquo;progressive.&dquo; The other pole of axis 1 was described by the terms &dquo;unethical,&dquo; &dquo;harmful,&dquo; &dquo;negative effects on welfare,&dquo; &dquo;personal worry,&dquo; &dquo;danger,&dquo;
and &dquo;inequality creation&dquo; (see Figure 1), reflecting negative attitudes. The
construct &dquo;tampering with nature/unnatural&dquo; was associated with the negative
constructs on axis 1 but was also differentiated by axis 2. Although &dquo;personal
objections&dquo; made an important contribution to axis 1, there was no differentiation in terms of positive or negative correlations with this axis, as the
correlation was positive for some people and negative for others.
was
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
107
Table 2.
Total Number of Constructs in Each Construct Class
and Number of Constructs with a High Loading on the
First Two Principal Axes (PA), Studies 1 and 2
(continued)
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
108
Table 2.
Continued
NOTE: To interpret the main perceptual dimensions of the spaces, the correlations
between each construct and each principal axis for each individual were examined.
These correlations provide a weighting for each attribute on a principal axis from which
it can then be decided which attributes are important. Constructs with a correlation of
~ -.51 or z .51 on the first two principal axes have been included. A construct can
correlate highly on more than one axis. Bold indicates that these constructs were
considered important determinants of attitudes toward the different genetic engineering
applications.
Study 2
The underlying constructs that determine attitudes toward specific applications of genetic engineering were elicited from twenty-five new respondents, fifteen females and ten males (mean age = 45.7 years, SD = ±16.25
years), from the Reading area, who were again paid £ 10 (approximately
U.S.$15) upon completing the experiment.
In this second experiment, the applications of genetic engineering were
described in very specific terms (see Table 1). Otherwise, the procedures, the
elicitation of constructs, and data analysis were the same as in study 1.
The construct classes are again listed in alphabetical order in Table 2. The
number of constructs elicited from each respondent ranged from 4 to 26, with
the mean number of constructs being 13.2. The constructs were allocated to
one of forty-one classes by two independent researchers, as in study 1. Figure 2
shows how the different applications were perceived in relation to the
constructs elicited from the respondents. The first two axes of the GPA group
average configuration together account for 71 percent of the variation in the
data. Axis 3 accounted for 7 percent, and subsequent axes 6 percent or less.
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
gp
1E~S
0) co 4)
’~ L
c ~’ojo.~
&dquo;z
f&dquo;+
~
<
o
t
i’’ ’0)>’0CCCof~ N
~
T3
~
>Q)(DM ~
’o
0
ccc
.2 0.2
0
(B c E
0
~ ~ T5
>§@
y@
~gCD
CL
c~ cuts
Q§§~
OJCTD
&horbar;
c
m
0
S
(Dcc> cc$-00
-C<0’(0
a) E
?rS sly
~gf&dquo;
w cm
C
x
~xgo
~ ~~
Q o£C
4m a)
O
I
a)
r-~ C
fi rL...
c
’a
0_N
0
-
I
C.C
~ NU
0.0 0
&dquo;t.BgS
c:
0c a
Sg.sjs
m C >
(D
o’§’o!5~>
0 0 0 Cl DM 2
t :2 -0 o
L
0 (D C T
m%,9E
2m&dquo;§fi
~ 3 ja S ’~
Co
pQ
co af
0L16o
z
r
v +~~. ~ a U 0 ~ ~U ~
N (D(~ a)~’ a)N
U)
0_ 0c
L-otmc
CL
- o-im C)
&reg; &reg; fl -° L N C ._ ?r U >
V
-~ ’a Q. ~ ’s o)
tcca
C U.~ 3 C
4D
0
-
O
0
E.02)CCo.en(~
uJS§-§
-
2
W
pf
Z c~a ~E ~
b.
li
.U
20
109
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
m N ~O CCE NO
&dquo;
~S~i~
~5o~~
;¡:: m;; ae .g1
r r- 0 0
Cl)c-Eas(l)
~ g ~,5 ~
:S4mg’ Q)Co<0 ~ c
r . 0 CO C a)
c~>.-c
S0
as ~~ 8.
m
.0
0 VD
CL
m
CI) m J2 (I) .CL
U)
~ii: 0x:6:g
Q) - (I)
-CJ
>
C~~~S
& ~ ~
co
S’5~~20
o 0
~S~Sj
C ~
41 l4 V > N
C1m
’~ TO+·
c 0 CL_
c
&reg;= d ~ N_ a0
8»
S~~
* - M (0 > !C:
.
Og<B-0<BS
g~~~s~
o ‘ N ,c~°a
.n m
.21:
~ë3ae.c
1âtm
’;
’(ij
5Le
3~ o»0 v~CM~0
g fi ~ gJ ~~
8:>O’CI!?SQ)
OM
(D
u’tJ.c~:5õ
;¡:: :g .6 (I) .9 (I)
ma
13S-’C
=,- 0 Em~mo
= (D .- 0 rM ~ E~ C (I) a.
0
10 tm 0. 0
e~Box~
0 a
(D
0 CD
N
~ CD
o. ’(ij.52 !&dquo;’o,c...
à);;
e
Z
mCmC3U
fi0
en as as :I: .2
c> È.!!1 g~
a
0a10 CM
CL
O ~- cc$
N2
c0
cL
’gmS o =O
g~
:E~g~’Q)
!£>2~§~
CD ’g.5 :ae0~
a
(D
0 Q) >co
C1 O
c
r-
t~~~j!
fti ~ N
~~
M
Q. C a.g.~:ae
~ °§W
QL ~ (0
mvv
(D
£
N2~Cl) ..(’110
.5~E~
~’ii
F- 0. c CD
Q)
co
~;
m
n110
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
E
0 0. a)
zto cm 0
111
Interpretation was therefore confined to the first two axes. Table 2 lists the
number of constructs that have high correlations with the first two axes.
The first axis explains 44 percent of the data variation. The terms used to
describe the genetic engineering applications on the negative attitude side of
principal axis 1 are &dquo;harmful,&dquo; &dquo;immoral/unethical,&dquo; &dquo;tampering with nature/
unnatural,&dquo; &dquo;long-term effects,&dquo; &dquo;negative effects on welfare,&dquo; &dquo;personal worry,&dquo;
and &dquo;risk.&dquo; The opposite side of axis 1 was described by the term &dquo;personal
objections&dquo; (see Figure 2). Positive constructs were described by the lower
side of axis 2 (&dquo;important,&dquo; &dquo;necessary,&dquo; and &dquo;advantageous&dquo;). The construct
&dquo;beneficial&dquo; described the opposite end of axis 1 to that which was associated
with negative constructs. &dquo;Beneficial&dquo; was also associated with the positive
constructs loading on axis 2.
Figures 1 and 2 show that respondents associated negative constructs
primarily with applications involving animals or human DNA. The concepts
of benefit and need was more strongly associated with applications involving
plants or microorganisms. Presenting applications of the technology in
concrete terms, in which tangible benefits are obvious to the respondent,
produces a more complex attitudinal structure. Comparison of the structure
of constructs associated with general applications (study 1) and the structure
associated with the specific applications (study 2) indicates that perceptions
of need may offset moral or risk-related concerns. Axis 2 in Figure 2 accounts
for much more of the variance than in Figure 1 and is explained by the concept
of perceived need for the technology. There are also indications in Figure 2
that some applications are seen as ethically acceptable or low in risk but also
unnecessary and not beneficial (vegetarian cheese, strawberries that grow in
frosty conditions). These &dquo;trivial&dquo; applications of the technology seem to be
linked to personal objections, indicating that perhaps individuals see no
societal problems with application but have what they think are personal
problems with these particular developments. These results were validated
in larger survey samples in study 3.
Study 3
Two questionnaires were constructed to assess public attitudes toward the
different applications of genetic engineering used in studies 1 and 2. Questionnaire A assessed attitudes toward the fifteen general applications used in
study 1, and Questionnaire B assessed attitudes toward the fifteen specific
applications used in study 2. The questionnaires assessed attitudes on each
of the psychological constructs identified as being important in studies 1
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
112
and 2. Construction of the questionnaire is summarized in Table 3. The two
respondent groups will hereafter be termed A and B, respectively.
In the repertory grid studies, the constructs &dquo;tampering with nature,&dquo;
&dquo;unnatural,&dquo; and &dquo;environmental effects&dquo; were difficult to separate within the
interview data as discrete constructs, as they tended to appear together.
However, to treat them as a single construct in the factor-analytic study was
inappropriate because linguistically they represented different constructs. To
avoid omitting key concerns, items assessing the three aspects independently
were constructed.
Similarly, it was difficult to separate the constructs &dquo;immoral&dquo; and &dquo;unethical.&dquo; Some respondents tended to use the two terms interchangeably,
whereas others appeared to differentiate between them. Again, items assessing the two constructs independently were included in the questionnaire, to
check if they were generally used in the same way.
Respondents were recruited by newspaper advertising in two local city
papers in the United Kingdom, Cardiff and Manchester. The advertisement
asked for &dquo;200 volunteers to help us with research concerning public opinions
about technologies and food production.&dquo; The wide socioeconomic and age
range of the sample is reasonably representative of the U.K. population.
However, the sample cannot be said to be truly representative of the population, as it is possible that respondents had stronger attitudes (whether positive
or negative) about food and technology than the U.K. population overall
(although not necessarily about genetic engineering in food production).
£3.00 (approximately U.S. $4.50) was paid following return of a completed
questionnaire. Two hundred respondents were recruited into each group, with
equal numbers from each city assigned to groups A and B. A total of 316
usable questionnaires (79 percent) were returned within the required time
period. Completed questionnaires were returned from 77 percent of respondents in group A and 82 percent of respondents in group B. Within group A,
64 percent of the respondents were female. The average age of respondents
in this group was 44.5, SD = ±15.4 years. Within group B, 60 percent of
respondents were female. The average age of respondents in group B was
47.8, SD = ±16.8 years.
Results
A principal components analysis of the mean scores of each attitude
characteristic in relation to each general application of genetic engineering
indicated a two-component solution (unrotated) accounting for 97 percent of
the variance. The first component was associated with &dquo;rejection factors&dquo;
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
Q)
Z3
.4c-c
i
0
r3
3M
0
il-I
0)
3
d
t
c
’C
dN
oc
0V
H
.a-
d
0
<c
0
0
d
13
a
:e
4
M
a
m
9
113
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
en
s
mQ)
-Z
c«ï
m
a.
,6
9)
?2
49
0a
c
a
c
m
’0
o
c
_m
8
*r.
C
0
C
’o
o
I,-
2
c
a.
m
M
a
H-
m
8
9
114
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
MJ
z
115
Table 4.
Loadings from the Principal Components Analysis
NOTE: Italics indicate variables loading
on a
principal component.
(accounting for 88 percent of the variance). Variables loading heavily on this
component were, on the negative attitude pole, &dquo;personal objections,&dquo; &dquo;immoral,&dquo; &dquo;unnatural,&dquo; &dquo;unethical,&dquo; &dquo;harmful,&dquo; &dquo;personal worry,&dquo; &dquo;negative
welfare effects,&dquo; &dquo;dangerous,&dquo; &dquo;risky,&dquo; &dquo;tampering with nature,&dquo; and &dquo;creation of inequalities&dquo;; and on the positive pole, &dquo;beneficial,&dquo; &dquo;advantageous,&dquo;
&dquo;necessary,&dquo; &dquo;progressive,&dquo; and &dquo;important.&dquo; This time, the second component (accounting for 9 percent of the variance) was labeled &dquo;long-term
effects,&dquo; as the variable of the same name loaded heavily on this component.
It should be noted that this variable also loaded on the negative attitude pole
of the first component. The third component accounted for less than 2 percent
of the variance and was not considered further. Component loadings are
shown in Table 4. A representation of the component space for the first two
components is provided in Figure 3.
Similarly, a principal components analysis of the mean scores of each
attitude characteristic in relation to each specific application of genetic
engineering indicated a two-component solution (unrotated) accounting for
98 percent of the variance. The first component was associated with &dquo;rejection factors&dquo; (accounting for 71 percent of the variance). Variables loading
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
M <C
_b/o
ii~.2
#s
Q.
8:~$
~2
c: Z ~
:EC.a
~c:’5
o<o-E
.n E v~
Q):::Ias
$#8
§Cgg
Bc~a’u
c5 &dquo;’ 5 c <c
,- as.-
ai g~i
o aec E.
u
S.~?o
%#.C3
~~-~
V§a~ ~ ~~C o
E
0, as.C8#
Q) âi
as:5 (I)
o
y /o*c
!o:-gc
,Q§~o
CI)
m vi ~
£s,f
c 1-:::1 a.
=-uii3as
@vf47 m rt..
Cm m-6.~’v)CD~
S 0. _ >
~
~~~
-a.c~õ
3
c
CI) Q) ’C C
~<e¡¡::Q)Q)
Ë]!-g
~
§’~§0)
m
0
c
s o a
CI) .- (I) Q)
S)S~-S
asasc:
to &horbar; t3
~8 c.2=:
-
0 M’XtO~:
SM
M
m
.!!2
g
a ~aev~ t
8: :6 u)~
as <C M M
fl
s
.
m~v~
~§-Sm
E ~~
¡_v~
á
-o0=L~
_0 i~i
aso.2
~8:5c:g>
c i’C âi
#£fi%
as c:
’&dquo;
u 0>’c:(1)c..:5 ’0
Ct
-
(I)
:is ’E .asas~
Q$
§«9
S LUJ-§
~~~i!
I- ~ g
ii: z.6 rJ
116
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
117
heavily on this component were &dquo;negative effects on welfare,&dquo; &dquo;personal objections,&dquo; &dquo;unethical,&dquo; &dquo;harmful,&dquo; &dquo;immoral,&dquo; &dquo;unnatural,&dquo; &dquo;risky,&dquo; &dquo;dangerous,&dquo;
&dquo;tampering with nature,&dquo; &dquo;creation of inequalities,&dquo; &dquo;personal worry,&dquo; and
&dquo;long-term effects.&dquo; The second component (accounting for 27 percent of the
variance) reflected more positive &dquo;acceptance factors.&dquo; Variables with substantial loadings on this component were &dquo;important,&dquo; &dquo;progressive,&dquo; &dquo;necessary,&dquo; and &dquo;advantageous.&dquo; &dquo;Beneficial&dquo; loaded on the negative pole of the
first component and the positive pole of the second component. &dquo;Long-term
effects&dquo; loaded onto both components. The third component accounted for
less than 1 percent of the variance and was not considered further. Component
loadings are shown in Table 4. A representation of the component space for
the first two components is provided in Figure 4.
Discussion
The data indicate that different applications of genetic engineering are
closely linked to perceptions of risk and benefit or need that are defined by
the nature of each application. It is unlikely that, for most members of the
public, attitudes toward the technology overall will define responses to
specific applications. In addition, increasing the specificity of application
types is likely to differentiate further public perceptions of risk and benefit.
Most negative attitudes are associated with genetic manipulations involving animals or human DNA. This confirms the results of previous research
and represents the most important determinant of attitudes toward the application of the technology. The public’s ethical concerns are associated with
applications involving the use of animals or of human genetic material, rather
than plants or microorganisms. The actual product of the application is less
important in defining these ethical concerns. The results imply that public
attitudes are defined by the processes associated with genetic engineering
rather than the products of these processes. &dquo;Unnaturalness&dquo; is also an
important determinant of underlying concern and again is focused on applications involving animals and human genetic material.
It has been argued that medical risks are seen as high in benefit and low
in risk, and thus acceptable, whereas nonmedical technologies are seen as
high in risk and low in benefit, and thus unacceptable (Slovic 1993). This is
clearly not the case here. The most important determinant of reaction appears
to be the use or involvement of animals or human genetic material, as opposed
plants or microorganisms.
The concept of need (or importance of benefit) has emerged as an important construct in the analysis. In the case of specific applications of genetic
to
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
N~~o
j!’°~g’s
m55’¡:
WvE ~c -aem
m§
c~ a~
c. ~ E .a~ c ° o-~-o o
&dquo;f1°§%$
~~
T
p~s~s~
COE
+:;Q):¡:¡Eas
C ~_V N U
~c:=_o
-SS’S.’Ss
’~~Q.~f~
~as.cQ).ae~
~ S
.~~.~
~
V
S.
M
cC c0 N O U_
Q)..!{2’E::::~
§S~2g(O C O C N
CI)5o ~:6~S~~§
as~~(1).6õ (I)~
~-5j3S3’~
E !~~=~
go ~§.Et:jD.$
Q)~.6~
~
Q ~ ~=&scaron; 8.
CCI) .~!
(I),5’-.s~~
N ~~ +~ ~ >
!
...
c
s ~s’s~&dquo;
~gS-gs’
?~iss~
y’~~~N
~
i
C
m
t~!!
õ°õ..c~
_ea.°’&dquo;
CCI)uI) (I)(cI) .c:E0=- Q)~ 0a.a.
i
?Õ
S ~ o <c <o
wC~pC
.gJ!!~:5
’C Q. C - .0~
Q) C)c:.aQ)>-,-’C :e
~Q.~ f!?~~i§
~g’5&eth;O
c :e ~
m c.5
& m°Q3/o9
(I)
co
mCl)Saso
(I) ~ :5 :: .6
SE &ofi=C
SS~jB~
as
%0
CI)8
.c c: Q) a.-=
s#qyfi9
CBlQ.~~~>
i~~
T5SS~3
8asi ~c8.°~
’C
O~ Q)m cCDm
?õ
t::
u J2.c~õ~
.9
.
~
~
8~!;
<C B I ~ ~0 ~
n
~ ~= Q
t-i&dquo;’a.
~ï5(uaeõ..E
m
~ !L~
~
J3’0jco:>.3
118
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
119
those that represented minor modifications to food products
associated with moderate concern but also tended to be rated as low in
terms of perceived advantages. It is likely, therefore, that public attitudes
associated with this type of application will be focused on questions relating
to perceptions of whether the technology is necessary (or whether these
applications are seen as &dquo;trivial&dquo;) rather than questions associated only with
perceived risk or ethical concern. Risk is more likely to be associated with
applications (related to food, agriculture, or medical) involving animals or
human genetic manipulation. Medical applications involving genetic transfers and using animals or human genetic material are also seen to be positive
in terms of importance or need, and their public acceptance may be determined by assessment of this &dquo;risk-benefit&dquo; tradeoff. Medical applications
(pharmaceutical development and genetic &dquo;screening&dquo;) that did not involve
animal modifications were seen as the most necessary applications, supporting the notion that the use of animals or human genetic material generates
the greatest objections.
In food-related modifications, perceived need was associated with agricultural applications in which modification appeared to facilitate production
or &dquo;process&dquo; rather than specific product characteristics. &dquo;Crops that can grow
in dry conditions&dquo; were seen as far more necessary than &dquo;strawberries that
can grow in frosty conditions,&dquo; perhaps because the product itself was seen
as a luxury item, and thus trivial.
Perceptions of long-term effects do not appear to be associated with either
acceptance or rejection. If negative perceptions are associated with a particular application, it seems that long-term effects will also be seen as negative.
If perceptions are positive, the long-term effects are considered acceptable.
The results appear to confirm the findings of previous research, that public
attitudes are likely to change when a gene transfer is tied to achieving a
specific goal, such as increasing nutritional content in a food crop (Powell
and Griffiths 1993), or when there are concrete implications of human health
(Heijs, Midden, and Drabbe 1993).
Ethical concerns appear to be greater for genetic engineering than for other
technological hazards (Slovic 1992) or other food-related hazards, for example, chemicals in food (Raats and Shepherd 1996). As important determinants
of the public reaction, ethical concerns and their potential impact on products
and applications in technology development must be carefully assessed.
The debate surrounding the technology should incorporate the concerns
of the public. Scientists should address ethical concerns directly both in their
research and when the results are disseminated to the public. Perceptions of
unnaturalness and questions regarding the need for the technology and its
potential benefits are as relevant in the public debate as issues of risk and
engineering,
were
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
120
The complexity of public attitudes suggests that the hierarchical
model of the dissemination of information described by Hilgartner (1990)
should be contextualized within a wider public debate that incorporates the
broader issues of concern to society.
Control of the technology did not emerge as a major concern in this study.
Control was not a determinant of consensus about different applications of
genetic engineering (whether general or specific), and it was rarely mentioned as a salient psychological construct in the repertory grid interviews.
Psychometric studies have found that control is an important characteristic
of risk perceptions for many hazards (Slovic 1987; Sparks and Shepherd
1994). Genetic engineering is characterized as a technology over which the
individual has very little personal control (Frewer, Shepherd, and Sparks
1994b); its control is seen to be the responsibility of science or the government (Frewer and Shepherd 1994). &dquo;Control&dquo; might not have emerged as an
important construct because the repertory grid method permitted the respondents to use alternative terms to describe it (Raats and Shepherd 1996).
Alternatively, other concerns (such as those related to ethical questions,
unnaturalness, and danger) may be considered more important, and the
corresponding constructs might therefore be generated with greater frequency. Control might also be a determinant of attitude toward all the
applications of genetic engineering, and thus the repertory grid method might
not elicit it as a differentiating construct.
Because risk perceptions are socially constructed, social interaction and
contact with others have been shown to affect attitude formations regarding
a potential hazard (Sapp, Harrod, and Zhao 1994). If the experimental work
reported here were repeated after group discussion, a different pattern of
results may be observed. Public acceptance of genetic engineering might
depend on both individual risk assessments and group assessments (those
derived from the public debate surrounding the technology and from discussion with other people). Collective public response may differ from individual responses (Blumer 1971). The importance of group decision making
should not be neglected, and further research addressing this important
potential influence on attitudes and subsequent decision making is required.
The question of individual choice also deserves to be investigated further.
Does the individual have any say in the development of, and exposure to, the
new technology? If, for example, food products of the technology are not
labeled, personal choice and control over consumption are severely limited.
Over and beyond this, however, public concern about ethical and environmental issues cannot be resolved through product selection and personal
choice alone. Unless the debate surrounding the technology is established
within the public domain, the opportunity for individuals to contribute to the
danger.
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
121
decision-making process surrounding the technology is
not a viable
option
(Joss and Durant 1995).
Our results also have implications for regulation surrounding the technology. For example, we have shown that the public is as at least as concerned
with the processes used by genetic engineering as with its end products.
Accordingly, the legislation regulating the technology should take account
of both areas of public concern.
Michael (1992) comments that the public differentiates between science
as an abstract entity (science in general) and as an activity directed at specific
events or problems (science in particular). Consideration of the future development of genetic engineering must consider public reactions to both abstract
and specific concerns, and address directly that which is most important to
the public.
Note
1. "Definition of genetic engineering: All living organisms contain a genetic ’blueprint’ which
required to produce the functioning cells. This information is contained
in the chemical DNA Genetic engineering is the transfer of parts of the genetic material from
one cell to another, or the introduction of changes into the genetic material. In this way, new
characteristics can be introduced into a cell. DNA may be transferred between different types of
cell, and it is possible to transfer DNA between microorganisms (like yeast or bacteria, for
example), plants and animals.
Genetic engineering allows the transfer of genetic material to occur between different living
things, so that the characteristics of the organism to which the material has been transferred can
be changed to create new characteristics which may be advantageous, or result in improvements
in production methods. There are many applications and potential applications of the technology
in the areas of food production, medicine, and agriculture. This experiment is about your attitudes
toward genetic engineering."
stores all the information
References
Advisory Committee on Science and Technology.
1990. Developments in biotechnology. London : HMSO.
Ajzen, I.1988. Attitudes, personality and behaviour. Milton Keynes, UK: Open University Press.
Arnold, G. M., and A. A. Williams. 1986. The use of generalised Procrustes techniques in sensory
analysis. In Statistical procedures for food research, edited by J. R. Piggott, 233-53. London:
Elsevier Applied Science.
Bauer, M. 1995. Resistance to new technology. Cambridge, UK: Cambridge University Press.
Blumer, H. 1971. Social problems as collective behaviour. Social Problems 18:298-305.
Dake, K., and A. Wildavsky. 1991. Individual differences in risk perception and risk-taking
preferences. In The analysis, communication and perception of risk, edited by B. J. Garrick
and W. C. Gekler, 15-24. New York: Plenum.
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
122
Dijksterhuis, G. B., and J. C. Gower. 1991. The interpretation of generalised Procrustes analysis
and allied methods. Food Quality and Preference 3:67-87.
Fischhoff, B., P. Slovic, S. Lichtenstein, S. Read, and B. Combs. 1978. How safe is safe enough?
A psychometric study of attitudes towards technological risks and benefits. Policy Sciences
9:127-52.
Flynn, J., P. Slovic, and C. K. Mertz. 1993. Decidedly different: Expert and public views of risks
from a radioactive waste repository. Risk Analysis 13:643-48.
Fox, M. W. 1988. Genetic engineering biotechnology: Animal welfare and environmental
concerns. Applied Animal Behaviour Science 20:83-94.
Frewer, L. J., C. Howard, and R. Shepherd. 1996. The influence of realistic product exposure
on attitudes towards genetic engineering of food. Food Quality and Preference 7:61-67.
Frewer, L. J., and R. Shepherd. 1994. Attributing information to different sources: Effects on the
perceived qualities of information, on the perceived relevance of information and effects on
attitude formation. Public Understanding of Science 3:385-401.
1995. Ethical concerns and risk perceptions associated with different applications of
genetic engineering: Interrelationships with the perceived need for regulation of the technology. Agriculture and Human Values 12:48-57.
Frewer, L. J., R. Shepherd, and P. Sparks. 1994a. Biotechnology and food production: Knowledge and perceived risk. British Food Journal 96:26-33.
1994b. The interrelationship between perceived knowledge, control and risk associated
with a range of food-related hazards targeted at the individual, other people and society.
Journal of Food Safety 14:19-40.
Gains, N., and M. H. Thompson. 1990. Sensory profiling of canned lager beers using consumers
in their own homes. Food Quality and Preference 2:39-47.
Gower, J. C. 1975. Generalised Procrustes analysis. Psychometrika 40:33-51.
Guy, C., J. R. Piggott, and S. Marie. 1989. Consumer profiling of scotch whisky. Food Quality
and Preference 1:69-73.
Hamstra, A. M. 1991. Biotechnology in foodstuffs: Towards a model of consumer acceptance.
In SWOKA research report 107. The Hague: Instituut voor Consumentenorderzoek.
1993. Consumer acceptance of food biotechnology: The relation between product
evaluation and acceptance. In SWOKA research report 137. The Hague: Instituut voor
&mdash;.
&mdash;.
&mdash;.
Consumentenorderzoek.
Harlander, S. K. 1991. Social, moral, and ethical issues in food biotechnology. Food Technology
(May): 152-59.
Heijs, W.J.M., C.J.H. Midden, and R.A.J. Drabbe. 1993. Biotechnology: Attitudes and influencing factors. Eindhoven: Eindhoven University of Technology.
Hilgartner, S. 1990. The dominant view of popularization: Conceptual problems, political uses.
Social Studies of Science 20:519-39.
Hoban, T. J., and P. A. Kendall. 1992. Consumer attitudes about the use of biotechnology in
agriculture and food production. Raleigh: North Carolina State University.
Hoban, T. J., E. Woodrum, and R. Czaja. 1992. Public opposition to genetic engineering. Rural
Sociology 57:476-93.
Jasanoff, S. 1993. Bridging the two cultures of risk analysis. Risk Analysis 13:123-29.
Joss, S., and J. Durant. 1995. The U.K. national consensus conference on plant biotechnology.
Public Understanding of Science 4:195-204.
Kelly, G. A. 1955. The psychology of personal constructs: A theory of personality. New York:
Norton.
Kraus, N., and P. Slovic. 1988. Taxonomic analysis of perceived risk: Modelling individual and
group perceptions within homogenous hazard domains. Risk Analysis 8:435-55.
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
123
Marlier, E. 1992. Eurobarometer 35.1: Opinions of Europeans on biotechnology in 1991. In
Biotechnology in public: A review of recent research, edited by J. Durant, 52-108 London:
Science Museum for the European Federation of Biotechnology.
Martin, S., and J. Tait. 1992. Attitudes of selected public groups in the U.K. to biotechnology.
In Biotechnology in public: A review of recent research, edited by J. Durant,109-42, London:
Science Museum for the European Federation of Biotechnology.
Michael, M. 1992. Lay discourses of science&mdash;Science in general, science in particular and self.
Science, Technology, & Human Values 17:313-33.
M. G., P. Slovic, I. Nair, D. Geisler, D. MacGregor, B. Fischhoff, D. Lincoln, and K.
Morgan,
Florig. 1985. Powerline frequency and magnetic fields: A pilot study of risk perception. Risk
Analysis 5:139-49.
Powell, D. A., and M. W. Griffiths. 1993. Public perceptions of agricultural biotechnology in
Canada. Proceedings, Food Technology (August): 14.
Raats, M. M., and R. Shepherd. 1991/1992. An evaluation of the use and perceived appropriateness of milk using the repertory grid method and the "item by use" appropriateness method.
Food Quality and Preference 3:89-100.
1996. Developing a subject derived terminology to describe perceptions of chemicals
in food. Risk Analysis 16:133-47.
Raats, M. M., P. Sparks, and S. Grugeon. 1994. Risk perception and the psychometric paradigm:
In search of new methods. Paper presented at the annual conference of the Society for Risk
Analysis, 4 December, Baltimore.
Sapp, S. G., W. J. Harrod, and L. J. Zhao. 1994. Social construction of consumer risk assessments.
Journal of Consumer Affairs and Home Economics 18:97-106.
Sharlin, H. I. 1989. Risk perception: Changing the terms of the debate. Journal of Hazardous
&mdash;.
Materials 21:261-72.
Sjoberg, L., and B. M. Drotzz-Sjoberg. 1994. Risk perception of nuclear waste: Experts and the
public. Report No. 16. Stockholm: Stockholm School of Economics.
Slovic, P. 1987. Perception of risk. Science 236:280-85.
1992. Perception of risk: Reflections on the psychometric paradigm. In Social theories
of risk, edited by D. Golding and S. Krimsky, 117-52. Westport, CT: Praeger.
1993. Perceived risk, trust and democracy. Risk Analysis 13:675-82.
Slovic, P., D. Macgregor, and N. N. Kraus. 1987. Perception of risk from automobile safety
defects. Accident Analysis and Prevention 19:359-73.
Sparks, P., and R. Shepherd. 1994. Public perceptions of the hazards associated with food
production and food consumption: An empirical study. Risk Analysis 14:79-86.
&mdash;. 1996. The moral dimension of attitudes towards genetic engineering in food production.
Unpublished manuscript.
Sparks, P., R. Shepherd, and L. J. Frewer. 1994. Gene technology, food production and public
opinion: A U.K. study. Agriculture and Human Values 11:19-28.
1995. Assessing and structuring attitudes towards the use of gene technology in food
production: The role of perceived ethical obligation. Basic and Applied Social Psychology
&mdash;.
&mdash;.
&mdash;.
16:267-85.
R. A. 1990. Media discourse and the social construction of risk. Social Problems
37:80-95.
Straughan, R. 1992. Ethics, morality, and crop biotechnology. Fernhurst: Surrey ICI Seeds.
TenBerge, J.M.F 1977. Orthogonal Procrustes rotation for two or more matrices. Psychometrika
42:267-76.
Stallings,
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016
124
Vlek, C., and P. J. Stallen. 1981. Judging risks and benefits in the small and large. Organizational
Behaviour and Human
Performance 38:235-71.
Zechendorf, B. 1994. What the public thinks about biotechnology. Bio/Technology 12 (Septem-
ber) :
870-75.
Lynn J. Frewer is a psychologist in the Consumer Sciences Department at the Institute
of Food Research in Reading, U.K. (Earley Gate, Whiteknights Road, Reading RG6
GB2). Her main current research interests are risk perception, effective communication
on risk attitudes, and the impact of scientific
of risk information, media influences
understanding on attitude formation.
Chaya Howard is a consumer scientist at the Institute of Food Research zn Readmg, U. K.
Her current research interests include understanding public attitudes to technology and
the role of social representation theory in attitude formation.
Richard Shepherd is Deputy Head of the Consumer Sciences Department at the Institute
of Food Research in Reading, U.K. He has worked extensively on the influences on
human food choice. His main current research interests are the application of psychological models of attitude and attitude change to food choice, the perception of risks
associated with foods, and risk communication.
Downloaded from sth.sagepub.com at PENNSYLVANIA STATE UNIV on March 3, 2016