Healthy Influence – Persuasion Blog

You might have encountered either Fooled By Randomness or the even more popular Black Swan, both written by Nassim Nicholas Taleb.  I just finished Randomness and see implications from this pop best seller for practical persuasion.

First, let me acknowledge my deeply felt dweebie nature and offer a professorial tut-tut about Taleb’s technical mastery:  He’s got some problems that would require a serious rewrite if these books were intended for a research crowd.  If you’re willing to throw yourself against those rocks, start with a search with the key term, “American Statistical Association.”

There.  I feel better.

Now to the good stuff.

Randomness is a slippery concept that has Alice in Wonderland properties, meaning that it means what the writer wants it to mean depending upon the circumstance.  Thus, randomness joins the blog pantheon of related concepts like Propaganda, Obscenity, and more lately, Strategic Communication.  You need to watch not only the words, but the meanings up the writer’s other sleeve to remain unfooled.

There are two major meanings of randomness.  The first is a highly technical procedural definition:  Randomness is the selection of one object from a population in a way that does not effect the selection of any other object in the population.  Random is not careless, haphazard, or whimsical.  It is a well known process that is observable not only in how you do it, but in what it produces.  Scientists use randomization as a crucial procedure when they select or assign as a means of equalizing samples of participants and materials and their interaction.  Randomization is a fundamental element in science.

The second meaning of randomness comes closer to the street meaning:  The mess of life.  There’s a lot of randomness in the stock market (an area of huge concern in Taleb’s books), for example.  In this sense random becomes synomymous with ignorance, error, and confusion.  Randomness in this sense is a major concern for observational fields like economics, epidemiology, climate studies, evolution and others that share one commonality:  You cannot randomly select or assign key features of the phenomena thereby bringing it under your control.  This lack of control in observational studies always makes it easier to get “fooled by randomness.”  However, and most oddly, you can use the known principles of randomness (see the first usage) as a means of understanding the mess of life.

Taleb rarely uses the term randomness in the first sense (an important procedure of scientific control) and almost always uses in that second sense of random as what we do not know in the mess of life.  You need to keep that difference in mind because Taleb offers many good illustrations of how people think they know something about a large complex system (like public health, global warming, or financial markets) when all they are doing is reading the tea leaves.

With these two different uses of the term, randomness, now clear, let’s consider the persuasion implications.  I see two.

First, most of what I write about is based on persuasion research that uses randomness in that first sense.  Researchers employ random selection and/or assignment as a means of controlling persuasion variables.  While one can still get fooled by randomness in this way, replication tends to point out the failure very rapidly.  That is, the first time I do an experiment, I may get bad lucky in my random selection or assignment and create biased outcomes that vary strictly as a function of my sampling or assignment and have nothing to do with the persuasion variable.  However, the next time someone replicates the experiment (and they will), randomization effects are more likely to be obvious as later studies fail to replicate findings.  Thus, the knowledge we generate about persuasion through this approach is more likely to be true (reliable, general, valid, useful) than research that never uses the approach.

Second, a lot of daily practical persuasion can be understood within Taleb’s point of view.  Practical persuasion usually occurs in a large complex social system of many variables that have a lot of randomness in their interaction.  We may get fooled by randomness when we achieve a desired outcome and think it was caused by our persuasion skill when it was really just a rare, unique, and chance combination of variables that no one understands and will probably never again recur.  If you never carefully test your persuasion principles against the Rock of Randomization, you will almost always be Fooled by Randomness.

In either case, randomization and its uses illustrates my Rule:

There’s a difference between persuasion, and smoke and mirrors; with persuasion the illusion persists.