Pathways to support genetically modified (GM) foods in South Korea: Deliberate reasoning,information shortcuts,and the role of formal education

Abstract

Analyzing survey data on the issue of GM foods in South Korea, this study examines two competing routes – deliberate reasoning versus information shortcuts – to forming opinions on controversial science. Findings indicated that both deliberate reasoning and information shortcuts were in play; but the process was moderated by a person’s education level. The well educated were more likely than the less educated to engage in deliberate reasoning when shaping their support for GM foods. Implications of the findings are discussed in detail.

Keywords

genetically modified foods public opinion science communication

1. Introduction

Although South Korea has imported genetically modified crops since 1995, the public only recently became aware that many everyday food items include genetically modified (GM) components (Park, 2008). While GM foods had not previously been the subject of substantial public discussion in South Korea, attention appears to have increased recently due to media reports suggesting links between genetic modification and health and environmental problems. South Koreans seem to support the use of genetic engineering, in principle. According to a recent national poll (Chang, 2008), about 63% of the respondents believed that GM organisms would benefit, rather than harm, consumers. At the same time, however, about 71% of the same respondents said that GM organisms might pose a serious health problem. A large majority (92%) agreed that the government should closely inspect and regulate GM crops.

How do people form their opinions on controversial science? Using the issue of GM foods in South Korea, this study examines two competing perspectives. One is to rely on deliberate reasoning (Kim et al., 2011). Many scientists have assumed that the public is able and willing to engage in informed and deliberate reasoning in order to make a valid judgment (Miller and Kimmel, 2001). Deliberate reasoning, in our definition, refers to a conscious and informed effort to learn about different attributes of an issue (such as the potential risks and benefits of GM foods), carefully weigh each, and take these all into consideration when forming an opinion. Many science communication researchers disagree and argue that the majority of the public should be described more accurately as “cognitive misers.” Instead of engaging in informed and deliberate reasoning, according to these researchers, average citizens tend to rely largely on available information shortcuts (heuristic tools such as political ideology, general trust in science, and religious beliefs) or information that fluctuates randomly when making a decision about a scientific issue (Cacciatore, Scheufele and Corley, 2011; Nisbet, 2005; Scheufele, 2006).

When thinking about a scientific controversy, do most citizens engage in a deliberate reasoning process? Or do they simply rely on available shortcuts? Using data collected from a professionally managed online survey panel, this study is one of the first attempts to explore these two competing perspectives in a single set of empirical analyses. While previous studies have often assumed that opinion formation relies on either deliberate reasoning (e.g., Miller and Kimmel, 2001) or information shortcuts (e.g., Cacciatore et al., 2011), we posit that forming a science opinion in large part involves a combination of both processes. More importantly, it is assumed that the extent to which a person relies on each process may depend on his or her cognitive sophistication, a concept often measured by level of formal education (Popkin, 1991). We assess the interaction between level of education and both deliberate reasoning and several different types of variables associated with the use of heuristic reasoning.

In order to fully understand how the public forms their opinions on science and technology, we also incorporate previous explanations from the science communication and the political science literatures (e.g., Brossard et al., 2009; Kim, Scheufele and Shanahan, 2005; Lee, Scheufele and Lewenstein, 2005; Popkin, 1991; Scheufele and Lewenstein, 2005). While these explanations can be applied to citizens worldwide, they have been tested mostly in the United States. South Korea offers an opportunity to examine whether the previous findings and theorizing can be applied to another country with different cultural and political traditions. Taken together, this study will offer an integrated view of science and public opinion in South Korea.

2. Pathways to support GM foods: Deliberate reasoning vs. information shortcuts

Deliberate reasoning is a form of systematic opinion formation (Eagly and Chaiken, 1993), where a person is well aware of different arguments made for or against a given issue, and carefully compares one with another when making a judgment. Not all but at least some of the population, particularly the so-called “issue public” for science, is likely to be well informed about potential risks and benefits of GM foods, and uses this knowledge in formulating an opinion. The notion of deliberate reasoning is consistent with the studies of “issue-specific” knowledge about new technology (Lee et al., 2005), and shares the assumptions that a) potential risks and benefits can be consciously calculated (Hornig, 1993) and that b) perceptions of risks and benefits are to some degree a function of knowledge about the topic (Wildavski and Dake, 1990).

Our deliberate reasoning model mirrors the ideal of an enlightened and informed public, a perspective that conceives of public opinion as being an outcome of conscious and knowledgeable deliberation (Nisbet, 2005). In this model, we consider perceptions of risks and benefits as an important part of the reasoning process. Learning about science often involves getting informed about certain risks and benefits of a given science or technology. We also believe that deliberate decision making necessarily involves weighing the risks and benefits of the alternatives. Social psychology theories that contrast the levels of cognitive elaboration, such as the Elaboration Likelihood Model of Persuasion (Petty and Cacioppo, 1986) and the Heuristic-Systematic Model of Information Processing (Chaiken, 1980; Eagly and Chaiken, 1993), have considered evaluating potential risks and benefits as an essential part of systematic opinion formation or the central – as opposed to peripheral – route to forming an opinion. Other cognitive approaches, such as the Theory of Reasoned Action (Fishbein and Ajzen, 1975) and Social Cognitive Theory (Bandura, 1989), have also incorporated the notion of beliefs about positive and negative outcomes in the process of making a decision or forming an attitude.

Contrary to the notion of an informed and enlightened citizenry, researchers in social psychology (Fiske and Taylor, 1991) and political science (Popkin, 1991) have suggested that average citizens are mostly “cognitive misers” who, whenever possible, try to minimize their effort to seek and process necessary information, and instead rely on a number of information shortcuts when formulating an opinion. What this means is that they tend to make a judgment with a little – and perhaps insufficient – information (Kunruether, 2001). In the same context, science communication researchers (e.g., Besley and Shanahan, 2005; Lee et al., 2005; Nisbet, 2005) have described a number of factors – such as ideological predispositions, trust in science in general, or religious beliefs – that individuals often use as decision-making shortcuts in order to reach a quick judgment about a science topic without engaging in extensive information seeking or deliberate reasoning.

Use of information shortcuts, however, should not be considered as entirely irrational. According to the bounded rationality perspective (Simon, 1991), the rationality of human decision making is limited by the amount of information, the cognitive ability, and the amount of available time. That is, because most people lack the ability and resources to arrive at an optimal decision, it is sensible for them to make a decision after having greatly simplified the choices available. Decision makers, therefore, tend to be a satisficer, one seeking a satisfactory decision rather than an optimal one. It is rational for individuals to try to minimize the amount of information costs necessary to make a decision (Downs, 1957); use of information shortcuts must be an effective strategy to reduce the cost of information and to increase the frugality of decision making. While decades of research have found that the opinions of average citizens are ill-informed (Delli Carpini and Keeter, 1996), minimally stable (Sniderman, Brody and Tetlock, 1991) and hardly consistent (Converse, 1964), the bounded rationality perspective suggests that the ill-informed opinions do not always lead to erroneous judgments. Rather, people can often reach an effective judgment despite the limitations because they usually have access to simple information cues that point toward an effective and valid decision (Popkin, 1991). In particular, in a sense that people will invest efforts in information seeking only when they find a reasonable pay-off, it makes good sense for them to rely on information shortcuts – rather than deliberate reasoning – in evaluating a science topic. For typical scientific issues – such as genetic modification – where gaining an in-depth understanding requires considerable efforts, the pay-offs in terms of being able to make a valid judgment may simply not be enough (Scheufele and Lewenstein, 2005).

While our deliberate reasoning and information shortcut models are both based on a cognitivist perspective, it is important to point out that understanding of decision-making processes should not be restricted to the individualist-cognitivist paradigm. The notion of social representation, for example, may offer an important alternative. The term social representation refers to the collective elaboration of a social object (such as scientific knowledge) by the community; it is depicted as both the process and the result of social construction (Moscovici, 1961). According to Bauer and Gaskell (1999), who applied the social representations theory to researching the public’s understanding of science, the transformation of scientific knowledge from experts to the mass public can be considered as a process of creative reconstruction, which involves associations, metaphors, images, and objectifications. These social representations of science, according to the researchers, mediate between the science world and the life world, shaping the public’s support for or opposition to a given science. To the mass public, popular science or scientific knowledge is an outcome of collective elaborations and constructions. In this regard, some science communication researchers have argued that potential risks and benefits of a technology are not necessarily objective facts; they are rather results of the interactions between scientists, different social groups (e.g., the media, interest groups), and individual members of the public (Brossard et al., 2009; Cacciatore et al., 2011; Scheufele and Lewenstein, 2005). The public’s evaluations of a specific risk or benefit may depend not only on their deliberate reasoning but also on how it is defined and framed in the public discourse.

Perceived risks and benefits of GM foods

Opponents and supporters of GM foods appear to base their opinions on several key risks and benefits. One of the major reasons to oppose GM foods is potential health risks when consumed (unsafe to eat). Although no substantial adverse health effects attributable to genetic engineering have been documented, many opponents claim that the potential long-term risks that GM foods can pose have not yet been adequately investigated (Key, Ma and Drake, 2008). Concerns about environmental and ecological risks are another reason some oppose GM foods (unsafe to the environment). Engineered crops may serve as conduits through which new genes spread to wild plants, which can then become weeds. The crops might also initiate a perturbation that may have unpredictable effects that ripple through an ecosystem (Mellon and Rissler, 2003). Opponents also point out that genetic modification can be seen as an act that violates a delicate natural order of things and thus leads to an uncontrollable disaster (uncontrollable disaster). “Frankenfood” is a metaphor that may well represent the danger of tampering with nature (Nisbet and Huge, 2006).

Meanwhile, supporters point to a number of agro-economical benefits. They argue that GM crops will bring higher yields and profitability to many farmers by helping them produce more, despite the odds or any environmental barriers (Raney, 2006). Another argument for GM crops is that they can lower pesticide usage (reduced pesticide usage). A recent study by the U.S. National Academy of Science (2010) reported that GM crops had in fact resulted in reduced pesticide usages and reduced soil erosion from tilling in the United States. Several scientists (e.g., Borlaug, 2000) have also argued that increased use of GM crops is probably the most feasible solution to the global food crisis and will help meet the rising demand for food, particularly in the developing world (solving world hunger).

In our study, deliberate reasoning refers to when a person is well aware of arguments about different risks and benefits, carefully weighs these and takes them all into account prior to expressing an opinion. In order to assess the extent to which a person engages in deliberate reasoning, we estimate how influential the potential risks and benefits are, as opposed to information shortcuts, in shaping opinions. More specifically, we examine how closely the six risks and benefits are altogether correlated with support for GM foods. The more closely correlated, we believe, the more likely it is that a person engages in informed and deliberate reasoning.

Previous studies have already well demonstrated that perceived risks and benefits are correlated with support for a specific science or technology (e.g., Brossard et al., 2009; Cacciatore et al., 2011; McComas, Besley and Yang, 2008; Siegrist, 2000). In our study, we are rather interested in exploring the nature of this relationship. Are perceived risks and benefits strongly correlated with support for GM foods? Or, do information shortcuts play a more significant role? Are the risks and benefits more influential among the highly educated? Or, are they equally influential across different segments of the population? We first test the following hypotheses:

H1a: Perceived risks will be negatively associated with support for GM foods.

H1b: Perceived benefits will be positively associated with support for GM foods.

Information shortcuts

While evaluating potential risks and benefits is a part of conscious and deliberate thought processes, a range of information shortcuts (or heuristic cues) can also be used to reduce the complexity of decision-making processes (Siegrist, 2000). In particular, we examine the role of ideological predispositions, religious beliefs, general trust in science, and opinion climate, the typical information shortcuts that have been tested in previous research on public opinion on controversial science (e.g., Brossard et al., 2009; Gaskell et al., 2005; Lee et al., 2005; Nisbet, 2005). In order to assess the extent to which a person relies on information shortcuts – as opposed to deliberate reasoning – we examine how closely these four shortcuts are associated with support for GM foods.

Ideology is the first potential shortcut we examine. As Nisbet (2005) points out, conservative and liberal members of the public may use ideology as a “selective screening device” for accepting or rejecting a new technology. For many, ideological predispositions can be used as a way to cut down on information costs (Popkin, 1991). Studies have indicated that conservatives are less likely than liberals to support embryonic stem cell research (Nisbet, 2005) and nanotechnology (Cacciatore et al., 2011). Religion is another shortcut. Religious individuals may find that science interferes with nature and is thus incompatible with their religious beliefs (Brossard et al., 2009). Studies have demonstrated that religious beliefs are an important source of opposition to controversial science and technology (e.g., Gaskell et al., 2005; Nisbet, 2005). General trust in science can be an important shortcut as well. A person’s overall trust in science, scientists and scientific institutions may allow him or her to make a judgment without understanding the nature of scientific debate (Brossard and Nisbet, 2006). Those who have a great deal of trust in science in general would be more likely than others to support emerging science and technologies (Lee et al., 2005). Finally, perceived opinion climate (or perceptions of other people’s opinions) may play a role. Science issues are often morally and emotionally loaded, and thus perceptions of what other people think (or which opinion is popular and which is not) may exert a significant social pressure (Collantes, 2008). Also, instead of carefully considering potential risks and benefits, people may simply use others’ opinions as a frame of reference when forming their own opinions on science (Collantes, 2008). The following hypotheses are put forward:

H2a: Being ideologically conservative will be negatively correlated with support for GM foods.

H2b: Self-identification with a specific religion will be negatively correlated with support for GM foods.

H2c: General trust in science will be positively correlated with support for GM foods.

H2d: Perceptions of favorable opinion climate will be positively correlated with support for GM foods.

Moderating role of formal education

Studies in political science have demonstrated that informed and deliberate political reasoning belongs mostly to well educated social elites, those who possess resources and cognitive skills to engage in such conscious decision making (Carmines and Stimson, 1980; Page and Shapiro, 1989). Education provides effective and efficient ways of gathering useful information. It also provides cognitive skills to be able to make sense of often complicated information (Eveland and Scheufele, 2000). Being highly educated and sophisticated, social elites share the ability and motivation to become well informed in the first place (Kim, 2008).

The same effect of education can be applied to making a judgment about a scientific controversy. That is, one’s use of deliberate reasoning and information shortcuts can be moderated by his or her level of formal education. It is likely that the better educated possess both greater motivation and greater ability to understand the science involved in such a controversy as genetic modification (Nisbet, 2005). The highly educated, therefore, would be more willing and able than the less educated to engage in deliberate reasoning. The less educated, on the other hand, are less informed and motivated, and thus more likely than the highly educated to rely on available information shortcuts. We test the following hypotheses:

H3a: Perceived risks and benefits will be more influential among the highly educated than among the less educated.

H3b: Information shortcuts will be more influential among the less educated than among the highly educated.

3. GM foods in South Korea

Although commercial planting of GM crops is banned by law in South Korea, the country has imported a significant amount of GM crops over the years. In 2010, for example, an estimated one million tons of GM soybeans and 1.1 million tons of GM corn were imported from a few different countries, including the U.S., Brazil, Argentina, and South Africa (Chung, 2011). As of 2010, 15 different GM crops – including soybeans, corn, tomatoes, and cotton – were being imported and distributed in South Korea.

With growing concern about potential health and ecological problems of GM foods, the issue has become highly visible over recent years. Media coverage has also increased considerably as the Korean government made several significant policy changes in recent years (Oh and Kim, 2011). In 2008, for example, the government approved, for the first time, the use of GM corn for human consumption, causing heated debates between the food industry, the government, consumers, and other interest groups. In response to the safety concerns, several new regulatory measures – including a new mandatory labeling law – were introduced and heavily discussed in the same year.

4. Methods

Data

Our survey data were collected in South Korea in July 2010, as part of the Korea Food and Drug Administration’s annual survey on the public’s perceptions of novel foods. Participants were recruited through a nationwide online survey firm from a panel of about 400,000 potential participants. Sampling was done via a non-probability quota sampling method, which included six age groups, 15–19 (n = 240), 20–29 (n = 240), 30–39 (n = 240), 40–49 (n = 240), 50–59 (n = 120), and 60 or older (n = 120), and an equal ratio of gender: males (n = 600), females (n = 600). Based on the quotas, email solicitations were sent out to the potential participants who met the criteria. Participants were then instructed that the survey was confidential and that they were free to withdraw from the survey at any time.¹ Individuals could not participate once their subgroup (age, gender) reached its quota. This means that a meaningful response rate could not be calculated. The purpose of our sampling design was to ensure representative diversity in terms of both gender and age groups.

Measurement

Respondents’ support for GM foods, the key dependent variable, was measured by asking how strongly (1 = strongly disagree; 5 = strongly agree) they agreed with each of the following four statements: “I often purchase GM foods” (M = 2.34, SD = .96), “I never consume any foods that contain GMOs” (M = 2.75, SD = .96, reverse coded), “I tend to purchase organic foods over GM foods although they often cost more” (M = 2.25, SD = .95, reverse coded), “I support use of genetic modification technology for producing foods” (M = 2.75, SD = .95). These four measures were later combined into a single index (α = .77).

Perceived risks and benefits were measured by asking respondents how strongly they agreed with each of the following six statements (1 = strongly disagree; 5 = strongly agree): “Potential environmental impacts of GM crops cannot be ignored” (unsafe to the environment, M = 3.54, SD = .99), “Genetic modification, by developing a new gene or a mutation, can become an uncontrollable disaster” (uncontrollable disaster, M = 3.71, SD = .95), “Most GM foods are unsafe to eat” (unsafe to eat, M = 3.57, SD = .86), “GM foods will enhance farmers’ income by innovating the way crops are grown” (agro-economical benefits, M = 3.14, SD = .98), “Genetic modification technology will help to reduce the use of pesticide” (reduced pesticide usage, M = 3.24, SD = .85), “Genetic modification technology would be the most effective solution to end world hunger” (solving world hunger, M = 2.85, SD = .99).

Ideology, our first measure of information shortcuts, was measured on a five-point scale (1 = very conservative, 5 = very liberal, M = 3.16, SD = .92). Religion was measured by asking whether the respondent was a Catholic (8.3%), Protestant (22.7%), or Buddhist (16.9%). General trust in science was measured with a total of eight items tapping respondents’ general belief in the contribution of science and technology, and their overall trust in scientists and scientific institutions (α = .73). Opinion climate was measured with three items asking how strongly (1 = not at all strongly, 5 = very strongly) respondents believed most of a) netizens (M = 2.43, SD = .71), b) family members and friends (M = 2.53, SD = .72), and c) Koreans (M = 2.51, SD = .71) supported GM foods. These three items were combined into a single index (α = .76).

Demographic controls included Age (median age between 30–39), Gender (50% female), Education (the highest degree completed, 49.1% some college or more), and Income (median household monthly income between 3–4 million Korean Won).

5. Findings

Our first two hypotheses (H1a and H1b) examine whether perceptions of potential risks and benefits are related to support for GM foods. We used a regression model with support for GM foods being the dependent variable (see Table 1). Six measures of perceived risks and benefits entered the regression along with information shortcuts and demographic controls (age, gender, education, income). Regression parameters were estimated using the ordinary least squares (OLS) method.

Table 1.

Regressions predicting support for GM foods.

	Whole sample	The highly educated	The less educated
	(N = 1200)	(n = 589)	(n = 611)
Demographics
Gender (female)	–.027	–.057*	.016
Age	.016	.039	.010
Education	–.020	–	–
Income	.012	–.003	.013
Incr. R2 (%)	1.6**	3.2***	.3
Information shortcuts
Ideology	–.009	–.007	–.007
Catholics	–.038#	.017	–.103**
Protestants	–.052*	–.057*	–.058#
Buddhists	–.012	–.002	–.031
Trust in science	.072**	.043	.090**
Opinion climate	.287***	.261***	.311***
Incr. R2 (%)	31.5***	33.5***	30.6***
Perceived risks and benefits
Unsafe to the environment	–.061**	–.015	–.096**
Uncontrollable disaster	–.185***	–.206***	–.159***
Unsafe to eat	–.315***	–.328***	–.308***
Agro–economic benefits	.105***	.159***	.064#
Reduced pesticide usage	.056*	.081**	.031
Solving world hunger	.092***	.068*	.108***
Incr. R2 (%)	20.9***	23.8***	17.9***
Total R2 (%)	54.0***	60.5***	48.8***

Note: Entries are standardized regression coefficients (final betas).***p < .001, ** p < .01, * p < .05, # p < .08.

Findings in Table 1 provide support for H1a (see the “Whole sample” column). As hypothesized, all three measures of perceived risks were negatively and significantly associated with the dependent variable: unsafe to the environment (β = –.061, p < .01), uncontrollable disaster (β = –.185, p < .001), unsafe to eat (β = –.315, p < .001). H1b was supported as well. All three measures of perceived benefits indicated a positive and statistically significant relationship to support for GM foods: agro-economical benefits (β = .105, p < .001), reduced pesticide usage (β = .056, p < .05), solving world hunger (β = .092, p < .001).

The next four hypotheses deal with the role of information shortcuts. Table 1 (see the “Whole sample” column) shows that ideology (β = -.009, p = ns.) is largely unrelated to support for GM foods. H2a was not, therefore, supported. When it comes to religion, being a Protestant had a significant negative relationship (β = -.052, p < .05), indicating that Protestants were less likely than others to support GM foods. Being a Catholic also had a marginally significant and negative relationship (β = -.038, p < .07). Being a Buddhist, however, was not related to the dependent variable (β = -.012, p = ns.). H2b was partially supported. Supporting H2c, trust in science also had a positive and significant relationship (β = .072, p < .01). It seems that those who have greater trust in science in general are more likely than others to support GM foods. Opinion climate was the strongest predictor of support for GM foods (β = .287, p < .001), suggesting that opinions of other people play an important role in shaping one’s own opinion on GM foods. This finding supported H2d.

The remaining two hypotheses (H3a and H3b) examine the moderating role of formal education. Are the highly educated more able and likely than the less educated to engage in deliberate reasoning? That is, are perceived risks and benefits more influential among the highly educated than among the less educated (H3a)? Are information shortcuts, on the other hand, more influential among the less educated than among the highly educated (H3b)? We first divided up the sample into two educational groups and applied the same regression model to these two groups of respondents separately: the highly educated (some college or more, n = 589), the less educated (high school or less, n = 611, see Table 1).

In order to test H3a, we examined how closely the six risks and benefits are altogether related with support for GM foods. The more closely related, the more influential the risks and benefits are, that is, the more likely it is for a person to engage in deliberate reasoning. More specifically, we estimated the amount of additional variance in the dependent variable explained by the six measures of risks and benefits in the regression (the R-square amount). This R-square estimate indicates how well or how much the six measures of risks and benefits can predict one’s support for GM foods. This shared variance, in other words, will demonstrate how important or influential the risks and benefits are altogether in shaping respondents’ support for or opposition to GM foods.

Table 1 shows that the perceived risks and benefits are more influential among the highly educated when forming support for GM foods. The six risks and benefits as a whole explained about 23.8% of the variance among the highly educated. Meanwhile, the same risks and benefits accounted for only about 17.9% among the less educated. One method to test the significance of the difference in R-squares obtained from separate samples (e.g., 23.8% vs. 17.9%) is to look at the confidence intervals (CIs) for the R-squares and see if they overlap. That is, by calculating a 95% CI for each estimated R-square, one can determine whether one R-square is significantly greater or smaller than the other. Each CI was obtained by computing the cumulative distribution function for the R-square (see Dattalo, 2008 for details). This particular method does not allow us to compare the magnitudes of R-squares estimated after control or other blocks of variables. All variables must enter the regression in a single block. The six risks and benefits, when entered in the regression without control and information shortcuts variables, explained about 54.2% of the variance among the highly educated (95% CI: 48.8% ≤ R² ≤ 59.6%). The same risks and benefits accounted for about 38.6% of the variance among the less educated (95% CI: 32.6% ≤ R² ≤ 44.6%). These confidence intervals do not overlap, indicating that the two R-squares (54.2% vs. 38.6%) are significantly different from each other. Taken together, these findings provide support for H3a. Perceived risks and benefits were considerably more influential among the highly educated than they were among the less educated. The highly educated seem to consider different risks and benefits of GM foods more carefully, suggesting that they are more likely to engage in deliberate reasoning.²

H3b was tested using the same method. The six measures of information shortcuts, when entered in the regression without control and risks and benefits variables, explained about 35.9% (95% CI: 29.9% ≤ R² ≤ 41.9%) of the variance among the highly educated. The same information shortcuts accounted for about 30.6% (95% CI: 24.6% ≤ R² ≤ 36.6%) of the variance among the less educated. These two CIs do overlap; that is, there is not a statistically significant difference in R-squares between the two groups. H3b was not supported.

The notion of the role of deliberate reasoning and information shortcuts being moderated by formal education was reinforced by the unique variance of each of the blocks accounted for (and not by any other variables in the regression model). The unique variance of each block was estimated by entering the block as the last step in a hierarchical regression. As shown in Table 2, perceived risks and benefits explained about 23.8% of unique variance among the highly educated, while information shortcuts accounted for only about 5.6%. Among the highly educated, in other words, perceived risks and benefits were about four times more influential than information shortcuts (23.8% vs. 5.6%). When it comes to the less educated, the ratio was only 17.9% vs. 10.2%, supporting the idea that in comparing information shortcuts to deliberate reasoning, the highly educated tend to rely more heavily than the less educated on deliberate reasoning.

Table 2.

Unique and shared variance in support for GM foods by education.

	The highly educated	The less educated
	(n = 589)	(n = 611)
Unique variance
Demographics	.5#	.1
Information shortcuts	5.6***	10.2***
Perceived risks and benefits	23.8***	17.9***
Shared variance	30.6***	20.6***

Note: Entries are R-squares (R²) in percent.

***

p < .001, ** p < .01, * p < .05, # p < .08.

6. Discussion

Analyzing survey data on the issue of GM foods, this study examined two competing routes – deliberate reasoning versus information shortcuts – to forming opinions on controversial science. We first examined deliberate reasoning by looking at how closely perceived risks and benefits were related with support for GM foods. As hypothesized, our six measures of risks and benefits were all significantly associated with support for GM foods. More importantly, we found that the extent to which a respondent engages in deliberate reasoning was moderated by his or her level of formal education. The six risks and benefits as a whole were more closely related with the dependent variable among the highly educated. This pattern of relationships suggests that the highly educated have thought about the issue more carefully and thus better understand how its potential consequences are related directly or indirectly to their own interests. The well educated seem more likely than others to engage in deliberate reasoning not only when evaluating a general political issue (Carmines and Stimson, 1980) but also when thinking about a scientific issue involving controversy.

While the highly educated and the less educated differed in their use of deliberate reasoning, no significant difference was found in terms of their overall reliance on information shortcuts. Of the four types of information shortcuts examined, religion, general trust in science, and opinion climate were significantly related with support for GM foods. While these information shortcuts have often been tested in the United States (e.g., Brossard et al., 2009; Cacciatore et al., 2011; Lee et al., 2005; Scheufele and Lewenstein, 2005), our study was one of the first attempts to explore the role of such shortcuts in South Korea. Our findings support the idea that the same information shortcuts can be applied to opinion formation not only in the United States but also in a country with different cultural and political traditions. It is possible, however, that the relative impact of each of the different cues could be different.

Among the information shortcuts examined, perceived opinion climate (or beliefs about the opinions of other people) had the strongest relationship with supporting GM foods, suggesting that people tend to refer to what other people say when making a judgment about a science topic. It is also possible that they simply conform to the opinion of other people (Fishbein and Ajzen, 1975). This finding, however, should be interpreted with caution as the same relationship can be produced by looking glass perception, i.e., people’s tendency to see others as holding the same opinions as their own (Fields and Schuman, 1976). This ambiguity of causal mechanisms reminds us not to over-interpret the influence of opinion climate estimated in our study.

Consistent with the findings in Europe and in the United States (Brossard et al., 2009; Gaskell et al., 2001; Nisbet, 2005), South Koreans’ religion seems to play a role in shaping their opinions on controversial science. The Catholics and Protestants in South Korea were less likely than others to support GM foods. Just like Christians elsewhere, the Korean Christians may perceive that genetic modification interferes with nature and is equivalent with playing God (Brossard et al., 2009). In this regard, it is interesting to note that being a Buddhist was largely unrelated, in comparison to the reference group of “no religion,” with either support for or opposition to GM foods. Buddhism, as it has evolved and as it exists today, does not recognize the existence of a Supreme Being or God as the Creator of nature.³

Overall trust in science, which tapped into our study respondents’ trust in scientists and scientific institutions, was also significantly related with support for GM foods. When scientists are perceived as the experts, and when scientific institutions are perceived as politically neutral and unbiased, the public may develop a sense of deference to scientific authority as a long-term socialized trait (Brossard and Shanahan, 2003). The public seems to apply their overall trust in science or their deference to scientific authority, as a form of information shortcut, to a wide range of scientific and technological controversies (Lee et al., 2005).

In our study, ideology was largely unrelated with South Koreans’ support for GM foods. As Nisbet (2005) points out, conceptualizing the role of public values in forming opinions on scientific controversies requires an understanding of the macro-level institutional arrangements that surround the controversies. In Europe, for example, individual-level “green” orientations have been linked to resistance to biotechnology, and these “green” opponents are for the most part politically left (Nielson, Jelsøe and Öhman, 2002). In the United States, on the other hand, resistance to biotechnology has been linked to religious traditionalists, who in large part tend to be Evangelical Protestants and conservative Catholics. It seems that such linkages and arrangements have not yet been established in South Korea.

Before further discussing our findings, it is necessary to discuss shortcomings of this study. In particular, it is important to point out that the data used for the study came from a national Internet survey based on a carefully constructed quota sampling. Regardless of the quality of the sample, however, we must question whether the use of an Internet survey limits the generalizability of the findings. The answer is probably “yes” given that certain subgroups of the population (e.g., the wealthy, the highly educated, urban residents) are more likely than their counterparts to use the Internet, and thus are over-represented in this type of Internet panel. This question, however, largely misses the point for two reasons. First, although a telephone survey might have provided a sample that better represents low socio-economic groups, telephone surveys are increasingly problematic. Specifically, such surveys must deal with a large number of unlisted telephone numbers, within-household randomization and, most importantly, the increasing number of cell-phone-only households. Second, the number of Internet users in South Korea has increased considerably over the years. According to recent statistics (Internet Statistics Information System, 2011), about 77.8% of the entire population used the Internet in 2010, and the number reached 99.9% when it came to people between the ages of 10 and 40. Also, inasmuch as the focus here is on exploring a pattern of relationships and not providing a description of specific views such as might appear in a newspaper, a quota-based approach seems adequate.

Another limitation concerns the problem of missing data. Because the respondents were allowed to withdraw from the survey at any time, it is important to question if there were any systematic patterns in terms of withdrawals or no returns, which may raise concern about potential bias in the sample. Given that the issue being examined (GM foods) may involve a complicated science, it is possible that people with less education were more likely to withdraw, making the sample somewhat over-represent the highly educated. It is also likely that the less educated were less willing to participate in the survey in the first place. Given that the highly educated were possibly over- represented in the sample and because they are supposedly more likely than the less educated to engage in deliberate reasoning, the overall effect of deliberate reasoning might be overestimated in our study, while the effects of information shortcuts were underestimated. Similarly, it is also likely that people with great concern about GM foods were more willing to participate in and complete the survey, while people with less concern were more likely to withdraw. Our sample, therefore, might have over-represented people with great concern. Given that concerned respondents are likely better informed about potential risks and benefits of GM foods, deliberate reasoning may play a greater role among these concerned respondents than among people with less concern. Again, it is possible that the role of deliberate reasoning might be somewhat overestimated in our study, while the roles of information shortcuts were underestimated.

With these and other shortcomings in mind, our study offers a meaningful analysis of science and public opinion in South Korea. It also speaks to the role of democratic processes discussed in much of the recent science communication literature. Whereas much of the previous literature questions the role of knowledge in views about science (e.g., Satterfield et al., 2009), our findings suggest that the highly educated better understand potential consequences of a controversial science and how these consequences relate to their own interests. More specifically, their judgments appear more likely to rely on deliberate reasoning than on other heuristics or the availability of a small number of random pieces of information. Evidence that many citizens are involved in informed and deliberate decision making is important in many respects. First, research (e.g., Rosenstone and Hansen, 1993) has consistently shown a meaningful link between the level of knowledge and civic engagement. That is, informed citizens are in general more likely to participate in the process of policy making. Not only the frequency of participation but its quality also matters. While random or uninformed opinions may cancel each other out and thus have little overall impact (Page and Shapiro, 1989), informed participation by those with real views seems likely to have a greater potential to shape policy in the interests of those participants. This means that the better educated are more able to articulate their interests, and this should therefore raise concern about potential unequal representation of the less educated citizens in the policy-making process. Though this is not a novel concern, the present results highlight an under-explored mechanism through which policies on science and technology may tend to serve the interests of the social elites. This influence could be further compounded in cases where many ordinary citizens view science as a matter pertinent only to scientists and therefore do not believe the public’s input is either important or necessary (Brossard and Shanahan, 2003).

This study sought to examine how citizens’ informed and deliberate evaluation of a scientific issue is elicited in the policy process. It is important to point out that both deliberate reasoning and information shortcuts are in play though the process is moderated by a person’s education level. Deliberate reasoning played a greater role among the highly educated, raising questions about popular sovereignty in the democratic processes in science. Developing a more in-depth understanding of these mechanisms is increasingly important, given chronically under-informed (Patterson, 2002) and polarized (Binder et al., 2009) citizens in different parts of the world.

Given the range of issues on the public agenda that involve questions about health and the environment – of which biotechnology is just one – it has never been more important for communication researchers to assess the dynamics underlying risk perceptions. Developing a more in-depth understanding of the degree to which risk and benefit perceptions have attributes that work alongside knowledge must be one important avenue for future research.

Footnotes

Acknowledgements

This work was supported by the Korea Food and Drug Administration (Project title: Investigation of risk perceptions about novel foods, grant number 10062-094).

Notes

Author biographies

Sei-Hill Kim is an associate professor in the School of Journalism and Mass Communications at the University of South Carolina. His research interests are at the intersection of the media and public issues.

Jeong-Nam Kim is an assistant professor in the Brian Lamb School of Communication at Purdue University. His research interests are in public relations and public opinion studies.

John C. Besley is an associate professor in the School of Journalism and Mass Communications at the University of South Carolina. His research explores the relationships between media use, citizen engagement and risk perceptions.

References

Bandura

(1989) Human agency in social cognitive theory. American Psychologist 44: 1175–1184.

Bauer

Gaskell

(1999) Towards a paradigm for research on social representations. Journal for the Theory of Social Behaviour 29: 163–186.

Besley

Shanahan

(2005) Media attention and exposure in relation to support for agricultural biotechnology. Science Communication 26: 347–367.

Binder

Dalrymple

Brossard

Scheufele

(2009) The soul of a polarized democracy: Talk, media, and attitude extremity during the 2004 presidential election. Communication Research 36: 315–340.

Borlaug

(2000) Ending world hunger: The promise of biotechnology and the threat of antiscience zealotry. Plant Physiology 124: 487–490.

Brossard

Nisbet

(2006) Deference to scientific authority among low information public: Understanding U.S. opinion on agricultural biotechnology. International Journal of Public Opinion Research 19: 24–52.

Brossard

Scheufele

Kim

Lewenstein

(2009) Religiosity as a perceptual filter: Examining processes of opinion formation about nanotechnology. Public Understanding of Science 18: 546–558.

Brossard

Shanahan

(2003) Do citizens want to have their say? Media, agricultural biotechnology, and authoritarian views of democratic processes in science. Mass Communication & Society 6: 291–312.

Cacciatore

Scheufele

Corley

(2011) From enabling technology to applications: The evolution of risk perceptions about nanotechnology. Public Understanding of Science 20(3): 385–404.

10.

Carmines

Stimson

(1980) The two faces of issue voting. American Political Science Review 74: 78–91.

11.

Chaiken

(1980) Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal of Personality and Social Psychology 39: 752–766.

12.

Chang

(2008) 2008 Biosafety white paper. Available at: http://www.biosafety.or.kr/bbs/mboard.asp?exec=view&strBoardID=bsn_009&intPage=1&intCategory=0&strSearchCategory=|s_name|s_subject|&strSearchWord=&intSeq=36241 (accessed 22 December 2010).

13.

Chung

(2011) 2011 Biosafety white paper. Available at: http://www.bioin.or.kr/board.do?num=217974&cmd=view&bid=system (accessed 8 February 2012).

14.

Collantes

(2008) The dimensions of policy debate over transportation energy: The case of hydrogen in the United States. Energy Policy 36: 1059–1073.

15.

Converse

(1964) The nature of belief systems in mass publics. In: Apter

(ed.) Ideology and Discontent. New York: Free Press, pp. 206–261.

16.

Dattalo

(2008) Determining Sample Size: Balancing Power, Precision, and Practicality. New York: Oxford University Press.

17.

Delli Carpini

Keeter

(1996) What Americans Know about Politics and Why it Matters. New Haven, CT: Yale University Press.

18.

Downs

(1957) An Economic Theory of Democracy. New York: Harper & Row.

19.

Eagly

Chaiken

(1993) The Psychology of Attitudes. New York: Harcourt Brace Jovanovich.

20.

Eveland

Jr Scheufele

(2000) Connecting news media use with gaps in knowledge and participation. Political Communication 17: 215–237.

21.

Fields

Schuman

(1976) Public beliefs about the beliefs of the public. Public Opinion Quarterly 40: 427–448.

22.

Fishbein

Ajzen

(1975) Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley.

23.

Fiske

Taylor

(1991) Social Cognition. New York: McGraw-Hill.

24.

Gaskell

Allum

Wagner

Nielsen

Jelsoe

Kohring

Bauer

(2001) In the public eye: Representation of biotechnology in Europe. In: Gaskell

Bauer

(eds) Biotechnology 1996–2000: The Years of Controversy. London, UK: Science Museum, pp. 53–79.

25.

Gaskell

Einsiedel

Hallman

Priest

Jackson

Olsthoorn

(2005) Social values and the governance of science. Policy Forum 310: 1908–1909.

26.

Hayes

Matthes

(2009) Computational procedures for probing interactions in OLS and logistic regression: SPSS and SAS implementations. Behavior Research Methods 41: 924–936.

27.

Hornig

(1993) Reading risk: Public responses to print media accounts of technological risk. Public Understanding of Science 2: 95–109.

28.

Internet Statistics Information System (2011) Usage statistics. Available at: http://isis.kisa.or.kr/sub02/?pageId=020200# (accessed 20 January 2011).

29.

Key

Drake

(2008) Genetically modified plants and human health. Journal of the Royal Society of Medicine 101: 290–298.

30.

Kim

S-H

(2008) Testing the knowledge gap hypothesis in South Korea: Traditional news media, the Internet, and political learning. International Journal of Public Opinion Research 20: 193–210.

31.

Kim

S-H

Scheufele

Shanahan

(2005) Who cares about the issues? Issue voting and the role of news media during the 2000 U.S. presidential election. Journal of Communication 55: 103–121.

32.

Kim

S-H

Scheufele

Shanahan

Choi

D-H

(2011) Deliberation in spite of controversy? News media and the public’s evaluation of a controversial issue in South Korea. Journalism & Mass Communication Quarterly 88: 320–336.

33.

Krosnick

Holbrook

Berent

Carson

Hanemann

Kopp

Mitchell

Presser

Ruud

Smith

Moody

Green

Conaway

(2002) The impact of “no opinion” response options on data quality: Non-attitude reduction or an invitation to satisfice? Public Opinion Quarterly 66: 371–403.

34.

Kunruether

(2001) Protective decisions: Fear or prudence. In: Hoch

Kunruether

Gubther

(eds) Wharton on Making Decisions. New York: John Wiley & Sons, pp. 259–272.

35.

Lee

C-J

Scheufele

Lewenstein

(2005) Public attitudes towards emerging technology: Examining the interactive effects of cognitions and affect on public attitudes toward nanotechnology. Science Communication 27: 240–267.

36.

McComas

Besley

Yang

(2008) Risky business: Perceived behavior of local scientists and community support for their research. Risk Analysis 28(6): 1539–1552.

37.

Mellon

Rissler

(2003) Environmental effects of genetically modified food crops – Recent experience. Paper presented to the conference, “Genetically Modified Foods – the American Experiences,” Copenhagen, Denmark, June.

38.

Miller

Kimmel

(2001) Biomedical Communications: Purposes, Audiences, and Strategies. New York: Academic Press.

39.

Moscovici

(1961) La psychanalyse, son image et son public. Paris: Presses Universitaires de France.

40.

Nielson

Jelsøe

Öhman

(2002) Traditional blue and modern green resistance. In: Bauer

Gaskell

(eds) Biotechnology: The Making of a Global Controversy. Cambridge, UK: Cambridge University Press, pp. 179–202.

41.

Nisbet

(2005) The competition for worldviews: Values, information, and public support for stem cell research. International Journal of Public Opinion Research 17: 90–112.

42.

Nisbet

Huge

(2006) Attention cycle and frames in the plant biotechnology debate: Managing power and participation through the press/policy connection. Harvard International Journal of Press/Politics 11: 3–40.

43.

Kim

(2011) Framing GM food in the Korean news media. Paper presented to the annual meeting of the Association for Education in Journalism and Mass Communication, St. Louis, MO, August.

44.

Page

Shapiro

(1989) Educating and manipulating the public. In: Margolis

Mauser

(eds) Manipulating Public Opinion. Pacific Grove, CA: Brooks/Cole, pp. 294–320.

45.

Park

M-S

(2008) Risks of GM food in the global food system. Environmental Sociology, ECO 12: 63–87.

46.

Patterson

(2002) The Vanishing Voter: Public Involvement in an Age of Uncertainty. New York: Alfred A. Knopf Publishers.

47.

Petty

Cacioppo

(1986) Communication and Persuasion: Central and Peripheral Routes to Attitude Change. New York: Springer-Verlag.

48.

Popkin

(1991) The Reasoning Voter: Communication and Persuasion in Presidential Campaigns. Chicago, IL: University of Chicago Press.

49.

Raney

(2006) Economic impact of transgenic crops in developing countries. Current Opinion in Biotechnology 17: 1–5.

50.

Rosenstone

Hansen

(1993) Mobilization, Participation, and Democracy in America. New York: Macmillan.

51.

Satterfield

Kandlikar

Beaudrie

CEH

Conti

Harthorn

(2009) Anticipating the perceived risk of nanotechnologies. Nature Nanotechnology 4: 752–758.

52.

Scheufele

(2006) Messages and heuristics: How audiences form attitudes about emerging technologies. In: Turney

(ed.) Emerging Science: Thoughts, Deeds, Analysis and Action. London: The Wellcome Trust, pp. 20–25.

53.

Scheufele

Lewenstein

(2005) The public and nanotechonology: How citizens make sense of emerging technology. Journal of Nanoparticle Research 7: 659–667.

54.

Siegrist

(2000) The influence of trust and perceptions of risks and benefits on the acceptance of gene technology. Risk Analysis 20: 195–203.

55.

Simon

(1991) Bounded rationality and organizational learning. Organization Science 2: 125–134.

56.

Sniderman

Brody

Tetlock

(1991) Reasoning and Choice: Explorations in Political Psychology. Cambridge, UK: Cambridge University Press.

57.

U.S. National Academy of Science (2010) The impact of genetically engineered crops on farm sustainability in the United States. Available at: http://www8.nationalacademies.org/onpinews/newsitem.aspx?RecordID=12804 (accessed 3 December 2010).

58.

Wildavski

Dake

(1990) Theories of risk perception: Who fears what and why? Daedalus 119: 41–50.