Exposure to Risk and Trust in Information; Implications for the Credibility of Risk Communication

The Australasian Journal of Disaster
and Trauma Studies
ISSN:  1174-4707
Volume : 2000-2

Exposure to Risk and Trust in Information;
Implications for the Credibility of Risk Communication

Britt-Marie Drottz-Sjöberg

Department of Psychology,
Norwegian University of Science and Technology,
NO-7491 Trondheim,
Norway

Abstract

The paper gives some examples of empirical studies where people have rated their experiences of risk, and of trust, in situations of increased risk exposure. Data are obtained from various studies conducted in Sweden, Russia, Ukraine, Belarus and Poland. They describe public reactions to radioactive contamination after the Chernobyl accident, environmental pollution, violence at work, and the prospect of siting a repository for high level radioactive waste in a community. Based on such heterogeneous examples, the paper aims at delineating systematic results showing that (a) the strength of the risk reactions is related to the factual severity of the hazard, thus disputing suggestions of rather arbitrary reactions to risk, and (b) that trust in information constitutes just one aspect of social trust. The latter phenomenon represents a more comprehensive concept which involves an integrated system of knowledge and personal experience. The relevance of the examples regarding emergency preparedness, disaster perception and especially risk communication is discussed.

Exposure to Risk and Trust in Information;
Implications for the Credibility of Risk Communication

Introduction

The identification of hazards plays a fundamental role in human life and for survival. Psychology has since long noted the uneven distribution of attention and weight given to negative or deviant cues relative to positive or confirming information. It is usually understood that negative information is more "diagnostic" (Fiske & Taylor, 1991). In a similar way we seem to weigh the credibility of information sources less if we have previous negative experiences of them than if we have no prior experiences. The connection we establish between an information source and the content of the information per se seems to be related to a validation problem. That is, if we are able to receive information, or to detect danger, with our own senses, or estimate risks on the basis of our own knowledge, we have independent means to establish facts or the truth. But if we do not, our judgement and reaction will have to rely on mediated information. Hence, the importance of trust in the information source.

A related aspect concerns how our fundamental ability to detect possible threats is affected when risk signals increase through an intensified bombardment of information, e.g. an information campaign or after an accident. The increased cognitive strain resulting from sorting impressions, processing information, and integrating or understanding its relevance is one factor to consider. Another is how we handle the positive and negative elements that we confront through the novel information. Increased information could result in increased general risk sensitivity, lead to exhaustion and fatigue, or depend on the processing capability in combination with the assessed "quality" and perceived relevance of the information. Psychological research has not firmly outlined the relationships between amount and quality of information, and individuals' reactions.

The aim of the paper is to contribute some findings and ideas to the discussion on risk experiences and the relationships between perceptions of risk, information, and social trust. All examples deal with exposure to externally generated, unwanted or negative hazards or events. The studies were conducted independently, and specific findings and similarities across studies are pointed out as being of potential relevance to the field of risk communication research. The examples provided here will suggest that perceptions of risk can be systematically understood, and generally support a positive relationship between the degree of exposure to a hazard and the experience of risk in that situation. It is also argued that there are many aspects applicable to credibility of and trust in information, and that such differences play important roles in people's understanding of, and reactions to, hazardous situations.

Findings from five empirical studies

Experiences from studies in Russia, Ukraine and Belarus regarding the Chernobyl accident

The data reported below were collected in two questionnaire studies within a collaborative project including researchers from the Commonwealth of Independent States (CIS) and the European Union (EU). The project, called the Joint Study Project 2 (JSP2) was financed by the European Commission and aimed at investigating the social and psychological effects of the Chernobyl accident. A pilot study in Russia in 1992 was followed by a major questionnaire survey comparing Russia, Ukraine and Belarus in 1993.

The pilot study (N=161) surveyed peoples' recollections of the Chernobyl accident, which occurred April 26, 1986 and their perceptions of the current situation. We had no access to census data or a control group for comparisons, but had some guidance in previous investigations (e.g. Archangelskaya, 1992). We wanted to get an idea of how people had received information about the accident, how they had reacted, and to what extent they trusted the information sources.

A sample of citizens in Novozybkov and the Bryansk area in the southwest of Russia, which was quite heavily contaminated by radioactive fall-out due to the accident, responded to a questionnaire. Most of them had received information about the accident during the week following the event, usually from TV, radio, via friends or by rumors. Thus, the information had usually been delayed for several days. Asked about the content of the first information about the Chernobyl accident, about 40% of the sample recalled learning about an explosion, about a third had got a message about a release of radiation, or heard about a fire at a nuclear power plant, respectively. They had believed the communicated content of this first information to a fairly high degree (a mean value of 4.9 on a 7-point scale). Thus, the results indicated credibility of this (negative) information. Asked about their trust in the primary information source, we found considerably more trust in information from foreign radio, local radio and local authorities than in information from friends and via rumors. The overall correlation between the belief in the conveyed information and trust in the information source was r=0.59. We could thus conclude that people differentiated between sources of information with respect to trust, and between the sources of information and their content. That is, there was fairly high credibility in the information that something dangerous had occurred, but more variance and generally less confidence with respect to the truthfulness of the reporting of the event.

Asked about their trust in 18 specified information sources with respect to information about the Chernobyl accident, the pilot study showed very low to little trust. As can be seen in Table 1, external or foreign sources of information were somewhat more trusted than "internal sources", although no information source was much trusted, i.e. the mean value reached slightly above the scale mid-point. (Scale: (1)="Do not trust at all", (4)="Neither much, nor little", to (7)="Trust completely"). These results all concerned the credibility of the information revealed about the accident, not if an accident had occurred. Thus, it might be feasible to clearly distinguish between credibility related to if something has happened and the credibility of various information sources regarding the reporting of aspects of a given event. (For details, see Drottz-Sjöberg, Rumyantseva, & Martyvshov, 1993).

The much more elaborate questionnaire study in 1993 included samples from Russia, Ukraine and Belarus (N=3067). We used a design (in each state) of a control group, a group of people living in contaminated areas (also called "restricted areas"), a group of people living in non-contaminated areas in close vicinity to highly contaminated areas, and a group of resettled people (i.e. people who had been moved from their previous home in a contaminated area). There were approximately 250 persons in each subgroup. The question of trust in various information sources was repeated. (For details, see Drottz-Sjöberg, Rumyantseva, Martyushov, Arkhangelskaya, Nyagu, & Ageeva, 1994).

There was very low trust in the listed information sources, especially in Ukraine. Trust in "external" information sources was again somewhat higher than for "internal" information sources, but no information source reached the scale level of "trust to some extent". We found a tendency among all groups of respondents to distrust information, in the sense that they believed that information about the consequences of the accident was withheld from the general public.

Table 1. Rank order of trusted information sources and mean values of ratings of personal trust in the information sources with respect to information about the Chernobyl accident among subjects of the pilot study, and separately for men and women.

However, when asked about whom they saw as responsible for taking care of those affected by the Chernobyl accident, most people marked "central authorities", in spite of the low trust ratings of national authorities. (See Table 2, where the numbers refer to type of area, i.e. 1= restricted, 2=non-restricted, 3= resettled, 4= control area). The results indicated that perceptions of responsibility, and thus trust in care taking and mitigation, had little to do with trust in information.

Table 2. Response percentages of who was seen as responsible for taking care of the people affected by the Chernobyl accident, by state and type of area.

We furthermore asked about people's perceived degree of exposure to "real risk", as well as their perception of "benefits", due to the accident. Although the latter question may be surprising, the risk mitigation effort after the accident could be evaluated in terms of benefits, due to e.g. drilling for water, paving of roads, etc. Approximately 30-40% of the respondents in the resettled areas across states believed that they had received a "dangerous radiation dose", and about 30-35% of these people believed that someone in the family had received a dangerous dose. Corresponding figures for the restricted areas were 16-37%; the non-restricted areas 7-18%, and the control areas between 2-11%. These results show clearly that respondents from the control areas perceived considerably less real risk than people in the restricted, and in the resettled, areas. It was of interest to note that people in the respective control groups perceived the same (small) amount of benefits that the respondents of the resettled areas.

Experiences from a study comparing Poland and Sweden regarding environmental concerns

This study from 1992 used a random sample in Sweden (N=1200, 18-65 years) and a sample in Poland (N=2001, 18 years and older). The response rates were 51 and 70, respectively. Its aim was to compare perceptions of environmental issues, especially related to industrial activities, wastes and pollution. We included ratings of risks and of benefits. One hypothesis, at this time in the early 1990's, was that people in Sweden were more informed about environmental hazards and therefore would rate risks higher than would people in Poland.

Mail questionnaires in both countries asked for estimations of personal risk regarding several everyday hazards to get "base-line data", e.g. drowning, to be hit by lightning, to have a heart attack, and to be assaulted. Furthermore, there were ratings of nine types of industries or activities, e.g. chemical-petrochemical-plastic industry, automobile industry, coal and nuclear industries. Each of these activities was rated from six different perspectives, e.g. the current risk, as well as benefit, of the activity for the respondent and his or her family, to the own country, and to one's country in ten years time. We also asked about the respondents' attitudes to each activity. Import of chemical wastes and radioactive wastes were the hazards that generally rendered the highest risks ratings. The respondents in Poland consistently estimated their personal risk higher than the Swedes, even with respect to being hit by lightning, and they consistently rated the personal risks related to the industrial activities higher than the Swedes did. The lowest risk ratings, which also displayed the smallest differences between countries, concerned the activities of small private enterprises and agriculture. Risks to one's own country were seen as higher than the corresponding personal risks among both Poles and Swedes. However, the differences between activities were larger than the differences between countries.

The results illustrate several points. First, that the perception of risks was more pronounced in Poland than in Sweden, contrary to the hypothesis that more environmental information would result in more concern in Sweden. The former group was, in the early 1990's, generally considered to be living in less favorable life conditions as compared to those in Sweden, including out-dated industrial facilities causing perceptible pollution, i.e. a more exposed group. Second, the difference between personal risk (here to "oneself and the family") and general risk (here to "one's country") could be documented also in Poland. And third, it was found that the public reactions to importing wastes were strongly negative in both countries. Ratings of especially the nuclear and chemical industries were also negative, although Poland did not have any nuclear reactors.

The findings seem to imply that people who live in a factually, and visibly, more polluted area but with less environmental information react more strongly to their situation than people who live in more affluent circumstances but have extensive information about environmental degradation. It is possible that increased exposure to hazards generally sensitizes people to risk, i.e. a possible interpretation of the overall higher ratings of everyday hazards, including being hit by lightning, as displayed by people in Poland (see also Sjöberg & Drottz-Sjöberg, 1991).

Experiences of hunter families in Sweden regarding radioactive fall-out

In 1994 hunter families were invited to participate in a study which included whole-body measurements of Caesium-137, a diary of food consumption for one week, and a questionnaire measuring personal habits, attitudes and reactions to the Chernobyl accident (Ågren et al., 1995). Respondents were selected from households where at least one adult member was a hunter, i.e. organized in a registered hunting team. The "hunter families" (N=286) were selected from areas with various degrees of radioactive contamination due to the Chernobyl accident (see Table 3). A control group of "hunter families" (N=183) was included in the study, selected from similar areas regarding radioactive contamination, but not measured for body content of Caesium-137.

Table 3. Respondents of the "measurement group"; county and local area in Sweden, mean deposition level of Cs-137 of the area, number of participants, and mean body content of Cs-137 (kBq) for men and women.

One reason to include a control group in the study was to investigate to what extent the personal invitation to whole-body measurements would affect the participants. The hypothesis was that people would experience such a suggestion as threatening or stressful. However, households in the "measurement group" participated to a much higher extent in the study than did households of the control sample, 90 and 72 percent respectively. We interpreted this outcome to signify that the invitation to whole-body measurement of radioactivity actually had motivated people to participate, i.e. they wanted facts about their actual exposure.

All subjects made ratings of risk for themselves as well as for others, for a number of different types of risks. Table 4 shows the results of such ratings regarding six sources of radiation. (Scale: 0="Non-existent risk" to 6= "Extremely high risk". For group categories, see Table 3 above). Possible injury due to radioactivity from nuclear test fall-out, and the Chernobyl accident, gained the highest risk ratings, and especially so in the most affected area in Gävleborg (group 3 in the table). Note that this group of subjects rated personal risk considerably higher than risk to "people in general" regarding the risk of injury due to radioactivity from the Chernobyl accident. This was an unexpected finding, possibly underlining the salience in these peoples' minds that they lived in the most affected area in Sweden. The table furthermore shows that the most exposed group, also more often rated other risks as higher, as compared to the other groups. This tendency could indicate a heightened risk sensitivity among those most affected, in accordance with the results of the Poland-Sweden comparison study, and of the more exposed populations in the JSP2-study.

Table 4. Mean values of ratings of personal risk and risk to "people in general" for six sources of radiation; hunter families in Sweden living in areas with various levels of ground contamination.

The differences between the ratings of personal risk and the rated ability to protect oneself from the risk, for each individual and source of radiation, were computed into a new variable of "controllability". And it could be shown that people found risks related to sunbathing, computer screens and radon as highly controllable, whereas the hazard of natural background radiation was perceived as less susceptible to personal protective measures. Especially radioactivity originating from nuclear weapons tests and from the Chernobyl accident were perceived as non-controllable. The results relating to the most exposed group in these respects exceed by far those from subjects living in less contaminated areas.

People in the most affected area of Gävleborg had found the information shortly after the accident in 1986 to be quite insufficient for their needs, whereas people in less affected areas were more satisfied with the information they had received. Regarding current (in 1994) information needs, those in the less affected areas were the most discontent. People who most often said that they really took the trouble of finding out the original place of production of the food products they bought lived in areas with a deposition level of approximately 40 kBq.m^-2, i.e. group 2. People in the most affected area, however, indicated instead consistently the need for measurements of radioactivity in food products in their community. Thus, the different experiences of exposure seemed to result in different reactions and demands.

In sum, hunter families living in the areas most contaminated by fall-out from the Chernobyl accident reacted the strongest. They rated their personal risk higher than the corresponding risk to people in general, for the specific and relevant hazard of injury due to radioactive fall-out from the accident. Contrary to our hypothesis, people seemed more motivated to participate in the study when the personal whole-body measurement was offered than when it was not. Those most affected by the accident also complained the most about the initial information, and they emphasized the need for measurements of food products in the community six years later rather than the need for information about the products' origin.

Experiences from a Swedish study on experiences of violence in the work-place

A study of perception of violence in the work place and society was conducted in 1993 and included a national sample of 1200 persons between 18 and 65 years of age (response rate 60 percent). There were 49 percent men among the respondents, and the overall mean age was 40 years (SD=12.7). An additional data collection involved 80 interviews, including risk rating tasks, with people working in jobs above average in vulnerability to threats and violence, i.e. taxi drivers and personnel in shops open late at night. The interviewees in each group were selected from both a large city and a small town. There was a majority of men in the taxi driver group and somewhat more women interviewed in shops.

The questionnaire for the national sample included a task where different professions were rated for risk of exposure to violent crime on 6-point scales. Participants perceived policemen, soldiers in the duty of the United Nations, social workers and taxi drivers most at risk among the categories presented. Also "shop cashiers" were rated above the "neither large, nor small" scale mid-point. These perceptions of vulnerability at the workplace, together with national statistics, supported our choice of interviewees.

Among respondents in the selected "vulnerable occupations", the taxi drivers gave a higher rating of personal risk on a 5-point scale of having a "traffic accident" than did the national sample (m=3.85 versus m=2.29), and rated their personal risk as similar to that of people generally. The taxi drivers also gave high personal risk ratings of "back injury" and "wear out" injury, "stress", and "physical violence" (m=3.33). They perceived the risk to themselves as very similar to the risk in the branch. Personnel in shops rated "heavy work", "back injury" and "wear out injury", as well as "stress" and being "overworked" especially highly. The mean rating of personal risk for "physical violence" in this group (m=2.18) was lower than their rating of that risk in the branch (m=2.80) and for people in general (m=2.86), but it was higher than the rating of personal risk in the national sample (1.87). These results clearly indicate that people react more strongly to those hazards that are salient in their personal situation, and that more exposed groups rate the risk to the relevant hazards higher than others do.

The examples from the studies above aimed at showing the consistency of the findings that more exposure to risk is related to an increased perception of risk, across very different types of studies. Suggestions that measurements of risk perceptions, or people's risk estimates, are fundamentally unreliable, or arbitrary expressions due to whim, emotionality or personal preference do not find support in these results. The final example below will touch upon a qualitative aspect of risk research and risk communication, i.e. the different ways people experience and interpret a phenomenon, and the possible underlying reasons for such different views.

Experiences after a local referendum regarding prospecting for a repository for high level nuclear waste in Storuman, Sweden

The total Swedish radioactive waste volume is estimated to about 200,000 m³ by the year 2010. Spent nuclear fuel and certain reactor parts will constitute about 25,000 m³ of the volume (SKB, 1996). In October 1992 the Swedish Nuclear Fuel and Waste Management Co (SKB) informed all Swedish communities about the development and program for deposition of high level radioactive materials and invited them to participate in overview studies. The community of Storuman was one of the few communities that responded positively to SKB's suggestion to conduct an investigation of its feasibility for possible future prospecting.

Storuman is located in northern Sweden, and has a population of approximately 7,500 people. The local government administration's invitation to SKB in late 1992 to conduct an overview study quickly resulted in opposition, and a "no to nuclear waste in Storuman" group was formed. People who were favorable to the invitation of SKB, or believed that the idea of future prospecting in the area at least should be investigated, followed by forming another action group. The local government continuously informed about the developments and results of the overview study, and organized a "reference group" to which local political parties and all active interest groups were invited. In February 1995 it decided to hold a local referendum in September of the same year regarding the future handling of the issue. The question put to the local public was if SKB should be allowed to continue the search for a location for a final repository in Storuman or not. Opinion polls related to the referendum had shown that 39 percent of those asked about their view had found the company to be trustworthy (71 percent of the "yes"-voters and 25 percent of the "no"-voters). More than half of the sample described the company as knowledgeable and sincere. However, only 28.5 percent voted "yes" in the referendum. This outcome points out that perceived sincerity and knowledge on the part of the company were not enough to convince the local public of the feasibility of further prospecting for a repository. The interviews were conducted to investigate the argumentation in more detail.

The results are based on interviews (Drottz-Sjöberg, 1996) conducted in the community of Storuman in November 1995. The 54 respondents were selected primarily among active members of political parties, the reference group, and interest groups, but also included persons in the local administration, business and industry. They commented on the overall development of the issue, and gave their interpretations to the overwhelming "no" (70.5%) outcome of the referendum.

The central themes resulting from the interviews were qualitatively categorized in a number of value dimensions. A dominant dimension of individualism (personal success) versus collectivism (solidarity) appeared. The opponents to a future repository usually emphasized solidarity, tradition, personal control, the need for a very high level of security, risk to future generations, and the importance of preserving nature or wilderness intact. Some of them underlined the need to develop small-scale establishments. Those who would permit the company to continue the search for a repository site instead emphasized the importance of (larger) investments and their beneficial spin-off effects, the need to sustain current and future employment levels, the possibility to influence the development, and the need for change.

Figure 1 shows a schematic representation of the elicited themes or values which the interviewees argued had influenced their voting behavior. It shows the major "lifestyle" dimension with the end points of individualism and collectivism, and the most often mentioned themes among the antagonists ("no"-voters) and protagonists ("yes"-voters). Thus, "individualists" were those favoring personal success and development, whereas "collectivists" emphasized the social group and the traditional ways of life in the community. The words placed on the left and right sides of the figure represent opposing points of views within the same "theme". For example, emphasizing "personal control" of developments contrasts arguing for the necessity of "major investments", and belonging to the original or "core group" of inhabitants of the community is opposed to having moved in from elsewhere. "Tradition" was highlighted against the greatly perceived need for development and "change", whereas requirements of "guaranteed safety" contrasted a view of guaranteed "influence" on the future development of the issue, or of a possible future repository.

Figure 1. Schematic structure of the interview responses ordered in relation to the "No" and "Yes" groups' argumentation and in relation to a main "life-style" dimension of "individualism" versus "solidarity".

The figure quadrants include the labels of four "interest groups" that emerged from the work. Towards the "collectivism" endpoint of the dimension the figure shows the "no"-attitude in combination with strong values favoring tradition, decentralization and guaranteed safety. This group is denoted "own housekeeping, village community". Experiences from earlier, historic times, traditional ways of life and a strong sense of belonging to an exploited and non-favored group characterized this group. The local government, which had invited SKB to conduct an overview study in the community, acted with the best interest of the community in mind. The group label "Communal investments" is therefore put close to the "collectivism" position of the life-style dimension. The intentions of the local government, however, were expressed in terms similar to the arguments raised by the "business" group, and this, together with the initiative to invite the company to the community, made some people see the local government as closely affiliated with the Yes-interests. Industry and business favored investments in their community, as well as expansion and the possibility to promote personal success, i.e. the core arguments in the "yes"-attitudes. But as can be seen in the figure there were also "no"-voters that favored some kind of individualism and personal success, here denoted "small, potential entrepreneurs". This group emphasized the possible future development of the community and region by tourism and diversified, small-scale business. It was however, in their point of view, not possible to combine the marketing of a clean nature and experiences of the wilderness with a repository for radioactive waste. . It was interesting to note that although democratic rules guided the entire development, there was a shift from representative democratic principles to direct public involvement rules over time. The majority of the population was "core group" inhabitants and "traditionalists" who favored their current way of life - their reason for staying in the region. It was hard to delineate a scenario including radioactive waste, which would accommodate this group.

As can be inferred from the figure, important values influencing the local referendum did not explicitly include the trustworthiness of the Swedish Nuclear Fuel and Waste Management Co, the information material, or the technical knowledge or expertise of handling radioactive materials. Respondents in the interview study instead underlined their experiences of respectable relationships, and sincerity of the company. Respect and sincerity, however, were not judged as directly relevant to the issue at hand, i.e. the future prospecting for a possible repository. Information had been abundant, but there had been a too strong emphasis on technical issues. There was too little done and understood about the social and psychological effects related to a possible repository, and its impact on social processes. However, although trust was not mentioned directly, the argumentation against the prospecting of a repository certainly included aspects of distrust. These sentiments were interwoven with accounts of conflicts of interest, e.g. the 20th century exploitation of the north, personal or family experiences of hardship related to living and laboring in a sparsely populated region of Sweden. There was pride and a positive survival attitude that comes with people who are used to solitude, waiting and the wilderness. They had not asked for nuclear power, the Chernobyl accident or an integration into the European community - in fact, they had already expressed their aversion to all of it. It was not relevant if the information about technicalities was trustworthy, they did not want anything to do with it.

Discussion

Risk perceptions have been measured in relation to a multitude of phenomena, as has the experience of trust. It has been shown that risk ratings differ systematically if the risk target is changed, i.e. relative to whom the risk applies (Sjöberg, 1991; Drottz-Sjöberg, 1993). The finding that people rate the risk to themselves as less than they estimate the same risk to others or to "people in general" has been interpreted as an "optimistic bias" (Weinstein, 1984), and related empirically to e.g. perceived control (Sjöberg & Drottz-Sjöberg, 1994). This paper has shown that when samples of exposed subjects or populations are asked about specific risks related to their life, the difference between personal risk and general risk may disappear – the personal risk estimate sometimes even exceeding that of others' risk. This was the case among people in areas especially exposed to radioactive fall-out after the Chernobyl accident, among people living in more heavily polluted areas, and among personnel in jobs of above average risk of being exposed to violent clients. We have elsewhere argued for the necessity of actually listening to arguments and to examine the factual circumstances before concluding that individuals or "the public" are ignorant or display mental health problems (Drottz-Sjöberg & Persson, 1993). A similar note of caution is due relative to arguments that outline risk experiences as solely socially construed and arbitrary expressions of recent influences. The implications for emergency preparedness is, similarly, to thoroughly investigate local circumstances before deciding on the optimal use of available resources. Risk communication after an accident meets with other demands than the task of public information regarding e.g. preventive actions (see e.g. Drottz-Sjöberg, 1998). The content must be tailored with respect to recipients as well as social and psychological circumstances.

The search for overall relationships between trust measures and public reactions to hazards seems futile in the here outlined framework, because needs, expectations and experiences differ widely due to both actual exposure and interest orientations of different subgroups. This is also why it is important to consider the positive relationship between exposure to risk and perception of risk, and to accommodate information, resources and e.g. mitigation efforts accordingly. Those who chose to disregard the relationship will be recommending, in the best case, ineffective risk mitigation efforts and risk communication strategies, and in the worst case, infuriate people and potential victims with their ignorance, and lack of consideration. Such mistakes may lay the foundation for social distrust. As the examples from the Chernobyl and Storuman studies indicated, trust in information is based on both what is addressed and what is left unanswered. There seem to be several kinds of trust, and trust in information or e.g. expertise seem to be facets of a larger, composite "social trust" concept, which involves collective memories or interpretations of historic events mixed with personal experiences vis-a-vis current developments.

Perceived personal control often counterbalances the estimated risk in the resulting perception of risk. It is an empirical question to what extent perceived "vulnerability" can be attributed to, or explained by, contextual weaknesses or psychological make-up. Such results would be of interest and of importance to risk communication research, and especially for the implementation of mitigating efforts after an accident. The lack of control, however, may be a more critical component in risk perception than the risk estimation element. Exactly the same type of risk information could be expected to have greatly different effects on populations who perceive a lack of controlling means, as compared to those who feel able to handle possible consequences. With respect to risk communication, the hypothesis implies that the communication of a fact may demand different information contexts to be similarly effective in different contexts. Regarding the communication of values, however, the situation is more complex since value systems are already deeply rooted in mental structures, in personalities and in cultures and sub-cultures. Thus, a lesson learned is that effective risk communication must be adapted to both the cultural context and to those concerned.

References

Archangelskaya, G. V. (1992). Population reaction to the radiation risk following the Chernobyl accident. In J. Baarli (Ed.), Proceedings of conference on "The radiological and radiation protection problems in Nordic regions". Oslo: Nordic Society for Radiation Protection.

Drottz-Sjöberg, B.-M., Rumyantseva, G. M., & Martyvshov, A. N. (1993). Perceived risks and risk attitudes in southern Russia in the aftermath of the Chernobyl accident. RHIZIKON: Risk Research Report No. 13. Stockholm: Center for Risk Research, Stockholm School of Economics.

Drottz-Sjöberg, B.-M., & Persson, L. (1993). Public reaction to radiation: fear, anxiety, or phobia? Health Physics, 64, 223-231.

Drottz-Sjöberg, B.-M. (1993). Risk perceptions related to varied frames of reference. P. Hubert and M. Poumadere (eds.), Proceedings of the Third conference of Society for Risk Analysis Europe. Paris: European Section of the Society for Risk Analysis, 1991.

Drottz-Sjöberg, B.-M., Rumyantseva, G. M., Martyushov, A. N., Arkhangelskaya, H. V., Nyagu, A., & Ageeva, L, A. (1994). Public reactions to the Chernobyl accident. Report from a data collection in 1993 in Russia, Belarus and Ukraine. Report in the JSP2 project. Stockholm: Center for Risk Research, Stockholm School of Economics.

Drottz-Sjöberg, B.-M. (1996). Stämningar i Storuman efter folkomröstningen om ett djupförvar {Sentiments in Storuman after the referendum regarding a repository}. SKB projektrapport PR D-96-004. Stockholm: Swedish Nuclear Fuel and Waste Management Co (SKB).

Drottz-Sjöberg, B.-M. (1998). Public reactions following the Chernobyl accident - implications for emergency procedures. Proceedings of the International Radiological Post-Emergency Response Issue Conference, Washington D.C., 9-11 September 1998.

Fiske, S. T., & Taylor, S. E. (1991). Social cognition. 2d edition. New York: McGraw-Hill.

Sjöberg, L., & Drottz-Sjöberg, B.-M. (1991). Knowledge and risk perception among nuclear power plant employees. Risk Analysis, 11, 607-618.

Sjöberg, L. (1991). AIDS; Riskuppfattning, attityder och kunskaper. En enkätunder-sökning av åldersgrupperna 30-45 år {AIDS: Risk perception, attitudes and knowledge. A questionnaire study of the age groups 30-45 years}. Rhizikon: Rapport No. 1. Stockholm: Center for Risk Research, Stockholm School of Economics.

Sjöberg, L., & Drottz-Sjöberg, B.-M. (1994). Risk perception. In Radiation and society: Comprehending radiation risk Vol 1. (pp. 29-59). Vienna: International Atomic Energy Agency.

SKB (1996). Djupförvar lokalisering. Förstudie Malå. Slutrapport {Repository localization. Prestudy Malå. Final report}. Stockholm: Swedish Nuclear Fuel and Waste Management Co.

Weinstein, N. D. (1984). Why it won't happen to me Perceptions of risk factors and susceptibility. Health Psychology, 3, 431-457.

Ågren, G., Drottz-Sjöberg, B.-M., Enander, A., Bergman, R., & Johansson, K. J. (1995). Radioactive Caesium in hunters and their families. FOA-R--95-00196-4.3--SE. Umeå: National Defence Research Establishment.

Copyright