Griggs: Experimental design, superstition and real success
Are you superstitious? Do you grab at anything that looks like it might help your department or your business? There is a better way to test what works. Instead of following fads or social media crazes, do what good researchers do to definitively show, at a certain level of confidence, what actually works. We want to find “causal relationships” — the actual drivers of results in the laboratory or success in your family, business or career.
If you want to become a truly great leader — stop with the platitudes and lean on science. To propel your company or department to greater success — update your understanding of experimental design. There are three variables every leader, manager or founder must know — the dependent variable; the independent variable; the confounding variable.
Dependent Variable — In experimental design, this is the outcome or result that is being observed. It is the end product you usually want to change or control. Other things are being manipulated while the researcher measures and observes the final outcome — the dependent variable. It is dependent upon something else. Examples include numbers of homeless on city streets, working capital in the bank or the amount of carbon in the atmosphere.
SPONSORED CONTENT
Exploring & expressing grief
Support groups and events, as well as creative therapies and professional counseling, are all ways in which Pathways supports individuals dealing with grief and loss.
Independent Variable — This is meant to be an isolated factor used to see what effect it might have on the final outcome (dependent variable). Researchers develop a prediction or hypothesis about the cause and effect (causal) relationship between what they’re manipulating (independent variable) and the final outcome (dependent variable). Examples include dollars spent on the homeless, increased sales incentives or emission screening and testing. Serious researchers are diligent at isolating this variable; business leaders are not.
Confounding Variable(s) — These tend to be extra events that occur during the experiment too often mistaken for having an impact on the final outcome. For instance, a small business wants to see if an advertising buy will have an impact on quarterly sales. During the same quarter a reporter publishes a glowing article on the business. The news article is a confounding variable. Sales may go up but you don’t know the real cause. You might say you don’t care which caused the increase in sales — but you do.
In research and in business, replication is crucial — can you demonstrate similar results again? If it’s dumb luck or a confounding variable, you won’t be able to control or replicate it the next time.
Superstitious Behavior — The legendary behavioral psychologist B.F. Skinner noticed that when he fed caged pigeons during an experiment, they would repeat whatever they had just done prior to getting the kernel of corn. If the birds had recently turned in a circle, they would turn another circle. If they had just pecked at the sky, that’s what they would repeat, connecting that behavior to more food. The famous “Skinner box” experiments thus documented the origins of superstitious behavior. “Temporal contiguity” is the term for when one behavior (turning circles) occurs closely in time to an event (getting fed). The events become associated with each other, whether or not one causes the other.
If your business strategy is built on wild hope and turning circles, you’re susceptible to superstitious behavior. If you’ve built a department or business on sound experimental science, you’ll focus more on isolated independent variables — to see what really works.
Rick Griggs is a former Intel Corp. training manager and inventor of the rolestorming creativity tool. He speaks on balance, teams, creativity and innovation. rick.griggs83@gmail.com or 970.690.7327.
