Perception of Risk the Ability

¶ … Perception of Risk

The ability to quantify and analyze the perceptual frameworks that groups, audiences and segments of people share in common, and then extrapolate those findings to define how the group of interest learns, perceives, avoids, and anticipates risks is one of the greatest benefits of psychometric research. Based on attitudinal data from primary research into specific audiences, groups, and constituencies, statistical routines including factor analysis, multi-dimensional scaling, data clustering, and structural equation modeling are used for defining the orthogonal nature of variables that report back perceptions, attitudes and beliefs regarding a specific topic. Psychographic research is completed using structured research designs where null and alternative hypotheses are defined, survey audience and sampling frame are defined, questionnaires are created, respondents contacted, and the results analyzed using statistical software including SPSS, SAS or BMDP. Each of these applications has the advanced statistical routines defined in this paragraph as options for analyzing data sets. It is important to note that the questionnaires developed to measure psychometrics rely on Likert scales and Myers-Briggs Type Indicator (MBTI) scaling which measures attitudes and behaviors on scales that are interval-scale in scope. This ensures reliability and validity of results, a critical part of the psychometric approach to measuring attitudes, beliefs, and perceptions. Myers-Briggs is a globally recognized industry standard for measuring attitudes, beliefs, perceptions and behavior and is often used for hiring, managing organizations, and planning how to create a high achieving culture in an organization (Myers-Briggs, 2006).

Slovic (1987) discusses the many implications of psychometrics in the perception of, reaction to and manipulating of fear as a specific strategy for modifying behavior. In the article the comparison of 30 factors with varying degrees of risk are tested with four groups including the League of Women Voters, a group of college students, active club members, and experts in the technologies associated with the list of 30 factors. The results of the psychometric testing indicate a very lopsided view of the perceived danger of nuclear power for example, with both the League of Women Voters and College Students saying it is the highest risk technology and the experts saying it is the least risky of all technologies.

In evaluating the validity and reliability of psychometric research in the context of Figure 2 from Slovic (1987) across the factors of Unknown vs. Known Risk (the vertical axis) and Controllable vs. Uncontrollable Risk (vertical axis) the variation in Caffeine's perceived risk vs. A nuclear war show the strength of using factor analysis to define the orthogonal aspects of risk when Myers-Briggs and Likert scales are applied to the data sets from the study's results presented. Caffeine is seen as a welcome stimulant to many of the millions of the world's workers every morning, yet nuclear war is seen as what must be averted to save the planet. Again, the strength of the methodology is shown in the ability to create a perceptual map shown in Figure 1 of the Slovic article. The distance between a hot cup of coffee and the unthinkable is used to show the contrasts possible when the orthogonal nuances of data are captured, analyzed, and presented.

Figure 1 also illustrates that when a risk is unknown and the implications of a negative consequence are high, both the factor analysis score of the risk attribute itself will increase from both an unknown risk and a uncontrollable risk. In effect, the greater the consequences and the greater the lack of knowledge, the most these two factors will correlate to drive the factor of interest into the upper right quadrant of Figure 1, where radioactive waste and nuclear reactor fallout.

Conversely when respondents see the risk as being controllable and to an extent observable, the factor of interest would inversely correlate to the knowledge of risk axis while clustering on the controllable vs. uncontrollable risk vertical axis, settling into the region of the quadrant where motorcycles, fireworks and alcohol accidents are. Again this is an approximation yet is the strength of psychometrics overall and the use of orthogonality to enforce validity and reliability is shown in this example.

In terms of communication of risk statement, it's clear from psychometrics that the more known a risk can become with credible, third-party data that accurately portrays its true threat in conjunction with an accurate measure of how controllable a risk is or not, equitability and individuality of risk, and the risk not being globally catastrophic would minimize risks as analyzed using the psychometric paradigm as defined in Figure 1 and throughout the Slovic article.

Finally under each of the following scenarios, a subject's perception of riskiness will increase or decrease is assessed.

Subject realizes that those exposed to the risk (including himself) may not know they are exposed to the risk

Realizing that with unknown risk the respondent would automatically be in the right two quadrants of Figure 1, and the severity of the risk would dictate where the respondent would finalize their perceptions, in the upper right or lower right quadrant. Given the fact that many people would over-analyze the fatalness of their exposure, most likely the event would be placed in the upper right quadrant of Figure 1.

Subject feels that while some experts might know the risk, s/he and many others like her do not know that they are exposed to the risk

Again this would push the respondents' perception to the upper right portion of the quadrant, due to the shock of learning they have been exposed and may have a permanent condition as a result of being exposed to a specific condition. The expert-level knowledge of would, in this situation, tend to not minimize risk as there is doubt of their complete and accurate knowledge of the risk itself. If the risk was bounded and well documented by the experts, the respondents' perception would drop to the lower quadrants.