Psychological and Socio-Cultural Theories of Risk

Psychological and Socio-Cultural Theories of Risk

Definition of Risk

The term "risk" is often defined differently depending on the particular paradigm. For example, risk is economics is typically defined in terms of differences in possible monetary outcomes and individuals/corporations involved in risk -- seeking behavior are typically seeking higher monetary payoffs (Markowitz 1952). When clinical psychologists, sociologists, law enforcement officials, and lay individuals identify "risky behaviors" they are referring to a broader meaning of the term "risk." In this context behaviors and involve risk are typically defined as behaviors that can be of potential harm to the person performing them or to other people (Steinberg 2008). In this sense the term "risk" is typically viewed in terms of possible negative outcomes as opposed to some other positive outcome such as the potential monetary gain.

This particular paper will assume that the definition of risky behavior includes some type of a dimension characterized by either a lack of control and/or potential disastrous consequences coupled with an unknown dimension that consists of unfamiliar, delayed, and/or unobservable consequences (Slovic 1987). In this sense risk is a property external to the person/organization that describes a fundamental indeterminism of potential future outcomes. Any decision-making occurring under conditions of risk would incorporate an evaluation of a probability distribution over the potential outcomes that include some type of judgment of the intrinsic value of each potential outcome to the person/organization.

Understanding the perception of risk and risk -- sensitivity is integral to psychological descriptions of decision-making and personality, understanding economic decisions, and understanding choice in terms of sociological and cultural variables (Rothschild and Stiglitz 1971; Tversky and Kahneman 1974; Loewenstein 2001). It would be difficult to discuss the notion of risk without discuss and the motion of choice as well which injects the idea of a volitional or directed selection of a desired action(s) motivated by personal preferences regarding specific outcomes. The challenges brought up by the notion of choice have been the result of major philosophical and scientific debates for centuries (e.g., Aristotle 1998). The notion of choice is generally envisioned as a type of conscious selection process; however, choices need not be the result of conscious decision-making processes as shall be discussed. Thus, risk relates to both the probability of an event happening and the consequences involved in the event and choice varies on whether people focus on the probability or the consequences of the event. It is with these definitions in mind that risk from both psychological and socio-cultural perspectives is discussed.

Psychological Theories of Risk

The earlier psychological theories of risk developed out of psychological models of decision making. The early research in this area is highlighted by Daniel Kahneman and Amos Tversky initially began studying how individuals approach gambling tasks and later developed a complex theory of risk of and heuristics and shortcuts in decision-making (see Kahneman 2011). The development of mental heuristics to aid in decision-making helps individuals conserve time and energy; however, the overuse of these heuristics can lead to severe errors in judgment and to what Kahneman and Tversky identified as cognitive biases (Tversky and Kahneman 1974). It is these cognitive biases that have defined much of the current psychological thought regarding decision-making and risk perception/assessment. For instance, Kahneman and Tversky first noticed that the students with whom they performed their research on were very quick to use the most available information to make decisions and determine risks and gambling tasks, even if the information led too poor choices. This led to their identification of the availability heuristic, which basically purports that information that can be easily brought into attentional focus is more likely to be used than information that is not easily brought into one's attention or is easily available (Kahneman 2011). However, due to such issues as the fallibility of memory or seeing things as being related when they are not (illusory correlation) the availability heuristic can become a cognitive bias and result in poor assessment of risk and poor decision-making.

Numerous other mental heuristics have been identified such as the affect heuristic, where people make judgments based on their feelings, is another such heuristic (Slovic et al. 2004) is another example. Overuse of this heuristic can lead to poor decision making and assessment of risk. For example Slovic found that people attributed high benefit and low risk to technologies that they liked and low benefit and high risk to technologies that they disliked (even in cases where there was overwhelming empirical evidence that the particular technology was not beneficial or it was actually dangerous).

Later as a result of numerous findings regarding heuristics and cognitive biases Kahneman and Tversky (1986) developed prospect theory as a method to ascertain how decisions and outcomes are ordered according to certain mental principles. According to prospect theory people edit and then evaluate information to determine risk. When editing people rank expected or potential outcomes according to whether there are benefits or detriments to them. Individuals then determine a reference point as to what would be acceptable and consider lesser outcomes losses and a greater outcomes gains. People have an aversion to loss an attempt to minimize losses over maximizing gains. Utility of the outcome is calculated as a reduction of risk aimed at keeping losses down. While this theory was their influential it was also found a poor number of issues with it and empirical research did not always support its predictions. For example, it was found that people overweigh information that they believe to be certain (certainty effect) and this can affect the assessment of risk in decision-making (Kahneman 2011). Secondly, the influence of demographic factors and societal influences is underplayed in prospect theory. The empirical evidence does not seem to indicate that everyone treats situations as a probabilistic event.

Psychometric research such as mentioned above has identified three major factors associated with how people perceive risk and how risk affects decision making: (1) the degree to which a risk is understood, (2) how much fear or anxiety it produces, and (3) how many people are at risk (the more people the more salient the risk; Kahneman 2011). However, this may not always occur via a conscious logical assessment that takes place in the form of an internal dialogue consisting of self-talk or an external dialogue. The more risky something is perceived the more visceral feelings anxiety and loss of control one experiences. Risks that are unknown evoke feelings of uncertainty. These can lead to implicit decision making.

Social cognition is a well accepted psychological paradigm that addresses how people interpret their environment, make decisions, and explain or justify their behavior (Schneider and Shiffrin 1977). The dual process theory of cognition states that cognitive processes involve in behavior operate on two levels: a slow and controlled process of thinking that allows for reflection and logical thought and a fast automatic implicit process that is generally unconscious and mechanized (Schneider and Shiffrin 1977). In order to develop or alter automatic thought patterns that are driven by procedural structures, schemas, and scripts one would have to invoke the controlled processes that are more prone to use logical, rationality, and reflection.

For example, Payne (2006) addresses a very important issue regarding how decision making and the perception of risk can be affected by implicit cognitive processes such as stereotyping. The inspiration for Payne's analysis is the shooting of a Ghanaian male by four Bronx police offers who mistook his reaching to his pants pocket as a gun threat. Thus, the question is, had the youth been white would the officers have fired on him? The "Weapon Bias" indicates that when having to make quick decisions regarding the presence of a weapon carried by a black or white target people significantly more often identify an object carried by a black person as a weapon. This bias has been demonstrated to emerge even when individuals are explained of its potential effect on their decisions and instructed to try and ignore it.

Payne (2006) explains that several factors drive the weapon bias: (1) a stereotypic factor that links African-Americans to violence and (2) the degree of intentional control people have over their responses. For instance, when not required to make a snap decision the bias is noticeably weaker. The prediction of the bias is best if one knows what the person's automatic response will be under conditions of little control and how likely is it that the person will not have control over their behavior? The key component in the automatic component of the decision is the automatic stereotype activation. There are several lines of evidence to support the bias from behavioral and neuroscience.

Behavioral evidence includes findings that people with more negative racial stereotypes are more likely to display the bias, and time to decide (which can override stereotypes). This appears to be mediated cognitive depletion. Neuroscience has investigated the association between two ERPs and the bias: the error-related negativity (ERN) associated with detecting conflicts between goals and cognitions, P200 associated with emotional reactions to threats, and N200 associated with cognitive control and conflict. Greater ERN and P200 activity demonstrate a stronger bias, whereas those with lesser N200 activity demonstrate it (Payne 2001). The key to reducing the bias is to either change the automatic impulse or maximize control. For example, officers with the most firearms training demonstrate the least effects of the bias. Using cognitive strategies to reduce implicit racial associations may also be useful.

But there are several questions that come to mind based on the Payne (2006) review. First, in his original study Payne used 31 college students, 24 women and seven men, all white. The targets were all male faces. So what is he really measuring in that study, racial bias of white women against black males? Payne's findings cannot be generalized at all. There are studies detecting strong in-group and out-group biases in African-Americans (e.g., see Hewstone, Rubin, and Willis 2002). If African-Americans demonstrate the same weapon bias as white females, how would this change the theory? Just what do implicit measures of stereotyping and prejudice actually measure.

For instance, Fazio and Olson (2003) reviewed implicit measures in psychology. There are a variety of techniques used such as priming (like Payne's research) where college students are generally quicker to identify distasteful or negatively charged words that have been primed with a picture of an African-American and are slower to identify positive words under the same conditions. Perhaps the most used measure is the Implicit Association Test (IAT), developed by Greenwald and associates (Greenwald, McGhee, and Schwartz 1998). This measure assesses the strength of an association between an attribute and a target concept by measuring the latency with which participants can employ two response keys that have each been given a dual meaning. So participants most often respond faster when negative stimuli are associated with African-Americans. Other measures include the completion of word fragments various GO/NO-GO procedures and others. There are also studies that use various physiological procedures. Interestingly, despite their sophistication, Fazio and Olson report that many of these studies are atheoretical in their assumptions and inconsistent in their findings.

For instance, Olson and Fazio (2003) used the IAT and a priming technique called the "bona fide pipeline" (BFP; Fazio, Jackson, Dunton, and Williams 1995) to measure racial stereotyping. The BFP assesses the evaluation activated in response to a prime by considering how the prime facilitates judging the connotation of following evaluative adjectives. While one would think that the IAT and BFP would show strong correlations, this is not always the case. In fact, Olson and Fazio believe that they access two different constructs. The IAT is based on the hypothesis that two categories associated in memory will be more easily represented by the same response key than when the categories are not associated. The BFP bases scores only on the evaluation automatically activated in response to an exemplar and this may or may not include category-level information.

Olson and Fazio had 92 white participants (61 females and 31 males) participate in an experiment with both the BFP and IAT. The BFP procedure involved four phases: In Phase 1, participants identified the connotation of 12 positive and 12 negative adjectives by pressing either a "good" or a "bad" key. In Phase 2, they saw Black, White, Asian, and Latino faces are were told to learn the faces and that would be tested later or to keep a mental tally of how many of the faces were of each ethnicity and that they would later be asked how many members of each race they saw (category condition). Phase 3 consisted of the test that participants anticipated they would complete. Phase 4 (the priming phase) combined the first two phases and consisted of four blocks. A prime, which participants were to either study or add to their racial tally, was presented briefly followed by a very brief interval and then the target adjective. Participants responded to the target as in Phase 1. At completion of the trials the participants in the category condition estimated the number of faces presented for one of the four races and participants in the traditional condition completed a face recognition test.

The IAT portion included 12 blocks of 50 trials. On any given trial, participants were presented with an exemplar of one of four categories: Black names, White names, pleasant words, and unpleasant words and categorized items by pressing one of two keys whose meanings changed depending on the block. Black and White names were categorized in Blocks 1 and 2, and pleasant and unpleasant items in Blocks 3 and 4. Blocks 5 through 7 were counterbalanced so that one of the races and pleasant words were assigned to one response key, and the other race and unpleasant words were assigned to the other key. Blocks 8 and 9 involved categorizing Black and White names, with the meaning of the keys reversed. Blocks 10 through 12 were identical to Blocks 5 through 7, but the race that was associated with pleasant items was now associated with unpleasant items and vice versa.

In the BFP findings participants significantly associated negative scores (stereotypes) in the category condition compared to the traditional condition. In the IAT results participants appeared prejudiced against Blacks and IAT scores did not vary as a function of BFP performance. The proportion of participants displaying some degree of negativity toward Blacks was significantly lower for the traditional BFP condition (about half) than either the category BFP or the IAT (over three quarters in both). The category BFP and the IAT were similar in this respect. These findings indicate that procedures that make race salient produce findings that often appear more prejudicial than those that do not. Thus context in these studies appears important. So in psychometric studies what is being measured?

Going back to Payne's inspiration for his research we need to ask ourselves were the officers acting on base-rate information or prejudice? Does gun training increase confidence or decrease prejudice? Likewise, in Payne's original study are young white female college students acting on the basis of unfamiliarity, culturally transmitted risk factors, or their own racist views? Would African-Americans act similarly in Payne's paradigm? These studies do not answer these questions. Another way to think about this is if a frog or mouse runs from objects such as sticks that resemble snakes are they prejudiced towards sticks or are they acting on the basis of evolutionary survival programs. In snap decision making where survival may be threatened false positives are more acceptable than misses. One miss and you are out. The research needs to consider the more complex nature of these relationships and the bias inherent in their measures. Psychometric studies of risk need to be fine-tuned and more research needs to be done regarding the conclusions of many psychometric studies.

Socio-cultural Theories of Risk

Where psychological theories of risk focus on the individual social -- cultural theories of risk focus on how group's, organizations, and society interact with the individual to confront the notion of risk. A popular socio-cultural theory of risk in the 1980s and 1990s was that of a Risk Society (Beck 1992; Giddens 1999). This was an attempt to describe the way in which modern societies evolve or organize in response to risk. The major issues involved in this theory were concerned with technological and environmental issues, thus the theory was criticized on these grounds. The major idea here was that as a society examines itself the changes itself in the process (reflexivity; Beck 1992). From a societal point-of-view is society can assess risks such as environmental pollution, crime etc. And reorganize itself in response to these.

For example, in previous non-industrialized societies the major risks were different (e.g., natural disasters) and society relied on institutions like the church to deal with risk. During the industrial revolution corporations became powerful and manual labor was important. People became less concerned with the church and more concerned with being loyal to corporations, business, and work. However, modern societies have been captivated by individualism as people became better educated and technology became available to everyone. This resulted in a highly educated information society which replaced the previous manual worker society. Instead of value being placed on loyalty to preview structures like corporations these individuals reflected on these relationships with these institutions and concluded that they were no longer necessary. Thus in the late 20th century there was a shift in personal meaning in Western societies from identifying with and being loyal to institutions to one in which the self was now defined as the primary agent of meaning (Giddens 1999). Thus, overnight the cultural meaning of risk and identity were shifted. One of the strengths of risk theory is that it offers an explanation as to how the shift in focus from "we to me" so to speak was affected modern malady; however, there is no reason to believe that technology caused people to shift this focus is more valid than the idea that a shifting focus of individualism drove the changes in technology and the desire for education. Moreover, risk theory focuses on environmental issues and technology and treats other social issues as rather secondary. Finally, risk theory explains risk perception as an individual quality that is in part shaped by environmental conditions. It is much more similar to a psychological explanation of learning than a socio-cultural theory.

The Cultural Theory of risk gained quite a bit of support differs from psychological theories of risk as well as risk theory in that instead of stressing perception and cognitive variables associated with risk Cultural Theory posits that the notion of risk is socially constructed. Cultural Theory attempts to explain how people and groups in society formed trust and mistrust, interpret danger, and calculate risk regarding their behaviors. Anthropologist Mary Douglas and political scientist Aaron Wildavsky are typically identified as the originators of this particular paradigm (Thompson, Ellis, and Wildavsky 1990). Decision-making is based on risk assessment/perception can result from three processes: (1) hard -- wired or instinctive behaviors that individuals automatically perform as a result of evolution or genetics, (2) behaviors that are culturally determined by our participation in a particular group or society and are subject to social norms, and (3) conscious decisions that occur when people feel they are exercising their free will or choice (Douglas and Wildavsky 1982). The last two of these are highly affected by societal norms and the first may develop as a result of evolution course planning to the tenets of the theory. Adherence to specific patterns of social relationships leads to distinctive ways of looking at the world or what Douglas and Thompson (Douglas and Wildavsky 1982; Thompson, Ellis, and Wildavsky 1990) referred to as cultural biases. According to this idea people tend to associate risk (in terms of its potential negative consequences) with behaviors that go against societal traditions. By defining risk in this manner, groups of individuals promote specific social structures and the process is therefore perpetual such that subscribing to specific patterns of social relationships influences how people view the world (these would be called cultural biases) and how people view the world reinforced or influence specific patterns of social relationships. Behaviors and viewpoints in a specific culture or across cultures are characterized along two dimensions that Douglas and Thompson termed "group" and the "grid" (Douglas and Wildavsky 1982; Thompson, Ellis, and Wildavsky 1990).