There has been a lot of research on the way mass opinion is formed. The research typically falls into two buckets: (1) diffusion models where opinion makers (elites) set informed opinion, which diffuses through the populace based on its persuasiveness and how supported the opinion is, and (2) interaction models where people hold independent opinions with varying levels of confidence, which are either reinforced or changed based on interactions with others.
There are variations on these two models that explore details, for example, the role of extremist views and clustering of like-minded people into opinion ghettoes. There is a growing body of work on the impact of Internet social networks on the process of opinion formation.
At root, this research is all about how to influence public opinion. It is instrumental. How can I, as an elite, leverage the process to achieve my political ends?
I would like to take a slightly different take on this. Why is it so many would-be opinion makers fail to do their homework and make so many errors of fact? Members of the media are theoretically supposed to do the homework before reporting, yet, errors creep in all the time. Is the error willful, laziness, the effects of time pressure, shrinking editorial resources, inability to detect “Astroturfing,” or that the Internet has made us stupid?
Willful Errors
An opinion maker (or informed agent who transmits information) willfully spread falsehoods or exaggerations in order to achieve some higher goal. Gun control is an issue area beset with claims on both sides that are factually inaccurate and uses ad hominem as a rhetorical technique.
Laziness
Presented with an issue, we formulate an opinion of that issue based on a combination of prior learning (stored knowledge) and through a process of categorization according to temperament and experience. The process is short circuited and we immediately “jump to a conclusion” in this model. It leads to error, because no effort was expended in investigation.
Time Pressure
In a 24-hour news cycle at Internet speed, opinion makers and informed agents in media operate under extreme time pressure which produces the same effect as laziness, although, not because the agent refuses to expend the time and energy in investigation, but because there isn’t time to do the investigation. The net effect is the same, bad information.
Shrinking Editorial Resources
I see this in my local paper. Basic errors that should have been caught by editors slip through to the printed page, and unlike the Web page, a printed paper cannot be corrected after the fact.
Inability to Detect Astroturfing
Organizations with agendas have weaponized the research on mass opinion formation and are very, very good at manufacturing opinion in a given issue area. This process is called Astroturfing (as opposed to authentic grass roots movements). Detection of Astroturfing is difficult. There are a few strategies, like hearing the same vocabulary from multiple sources, the use of ad hominem (attacking the characteristics of the opponent), and genetic arguments (attacking the source of the argument). The way Astroturfing works is to make an opinion seem more widespread, and therefore more influential, than it is. Make it seem to be generally held common sense and people will be more persuaded.
The Internet Makes Us Stupid
The Internet is a tremendous store of data and information. The Millennials and the iGen are the first to have grown up with instant access to huge stores of information available at a key stroke. Acquisition of knowledge requires effort. Reading at that level is work. Since there is so much information available for recall, it is merely practical that a person would just use Google as memory. The problem with this approach is that in the process of knowledge acquisition, a person learns a set of skills to evaluate information presented: such as logic and rhetoric for fallacy detection, synthesis through typologies and analogies, use of statistics, and experimental design. Without the effort expended in knowledge acquisition a person is crippled in the ability to form informed opinion, and temperament takes over.
In an ideal world, a person presented with an issue would first search his/her existing store of knowledge and, having a motivation to learn, would conduct a search of existing information, filter that information based on the merits, while using the meta-self (self-criticism) to assess personal bias, coming to an informed conclusion.
Instead what seems to happen is one of two methods:
First, when presented with an issue, a person searches the existing store of knowledge, finding none, merely formulates an uninformed opinion or accepts an argument from authority as long as it conforms with the person’s temperament.
Or there may be a failure to apply a bias failure if there is indeed a desire to learn.
In this era of “alternative facts,” “fake news,” Astroturfing, and opinion ghettoes, how can we as a citizenry address this problem? Do the work is one. For example, it has been reliably shown that handwriting notes is superior to typing. Learn how to read a book. Bring back the study of rhetoric. Second, be curious and skeptical, but temper skepticism with common sense. Radical skepticism, the root of academic schools of thought labeled Critical and Theory, with a capital C or T, does not lead to knowledge but destroys it. It is corrosive to positive knowledge. It is solipsistic. When I take the epistemological position that only what is in my own mind is “real” and there is no objective reality, then “alternative facts” are no longer “fake.”
Thomas Reid, a philosopher in the Scottish common sense tradition, wrote: “It is so irksome to reason with those who deny first principles, that wise men commonly decline it.” Some skeptical criticism of positivism (the philosophical position that positive objective knowledge is possible) is valid. Positivism cannot answer every question—e.g., metaphysical questions are a misapplication of positivism—but it can answer many questions. Radical skepticism denies the ability to gain positive knowledge on first principles. However, radical skepticism fails the lamp post test. Were we to take the position that the senses are unreliable and therefore no reality exists outside my own mind, then we would forever be walking into lamp posts, since they are unreal. Advice: Be skeptical of the radical skeptics and decline to adopt their first principles.
A Final Note
There is no denying the challenge of knowledge specialization. So much new knowledge is produced in narrow specialized areas that the average person is not capable of synthesizing it due to lack of context and knowledge. Based on that fact, a person of average intelligence is now required to accept arguments from authority. However, the person of average intelligence can be armed with certain tools for evaluation. For example, publication bias is a real phenomenon. Only surprising findings make it into print. Scientific confirmation of common sense is viewed as proving the apodictic and never makes it into print. Furthermore, negative findings never make it into print, so do not assume that one study makes an established fact.
Be extremely suspicious of epidemiological studies. Unless there is miraculously a “natural experiment,” where all other plausible causal factors except the one under investigation are controlled for, these studies are highly suspect. They rely typically on self-reported data, possibly non-random samples, and assumptions of population homogeneity. Epidemiological studies are the beginning of the formulation of a research question, not the end. Correlation is not necessarily causation. Without controlled experiment and establishing a causal chain, findings are an unconfirmed hypothesis as to causation. This is why government diet advice flip-flops so much—don’t eat too many eggs, because they cause heart disease—oops, my bad—you can eat eggs again.
In social science, do not make the assumption that an empirical finding is reliable. Social phenomena are complex and any study applies a theoretical approach that defines the model (a simplification of reality) to explain a phenomenon. The theoretical approach defines what variables will be omitted in a study. Correlations without a strong argument for causation are just that, a correlation. The model chosen can determine the result. Take for example the debate over IQ and life outcomes. When statistical regression is chosen as a tool for analysis, you’ve just made the assumption that factors do not interact, unless you create interaction terms a priori. The variables are independent causal factors that are pitted against one another for statistical significance according to an arbitrary threshold (the conventional test is a less than 5% chance that the result is in fact accident). Thus, there are two sides of the debate: People doing parsimonious (few variables) regression models who say that IQ and not upbringing is highly determinative of life outcomes (the quantitative researchers) and people who take an “intersectional” approach that say that multiple causal factors interact in non-linear and possibly incommensurable, ways to produce the measured effect (the qualitative researchers).