The Statistical Concept of ‘Regression to the Mean’ – and How It Impacts Your Laundry Operation
I’m presently working toward a doctorate degree in clinical psychology. And, as part of the program, I’ve been required to take a number of courses in advanced statistics.
Although psychology and stats may not seem compatible at first, statistics enables us to infer a number of things about the behavior patterns of practically anything being investigated by using small samples.
Therefore, in this column, I’ll explain how to use a few statistical principles to help analyze the behavior patterns of your laundromat business and to account for variations in sales, for example, at various times during the year.
It’s far more scientific than simply guessing or philosophizing why variations in your sales occur. And, if used properly, statistical analysis can significantly affect the accuracy your business decisions in many ways.
The specific statistical concept I’m referring to is technically called “regression to the mean,” and it’s far less complicated than it sounds.
Learning to recognize when regression to the mean is in play can help you avoid misinterpreting data and seeing patterns that don’t really exist – and, in turn, making poor and costly business decisions because of it.
To successfully operate your laundromat business, it’s important to minimize instances of bad judgement and address the weak spots in your reasoning. Learning about regression to the mean will help you do just that.
This notion was first developed by Sir Francis Galton (1822-1911), an English Victorian-era statistician, sociologist, psychologist, anthropologist, tropical explorer, geographer, inventor, meteorologist, geneticist and psychometrician. If that’s not enough, he’s also the guy who invented the very first weather map.
The idea of regression to the mean simply states with any complex events that are dependent upon many variables – where chance is involved (such as your business volume for a specific period) – extreme outcomes tend to be followed by more moderate ones.
The infamous Sports Illustrated jinx is an excellent example of this phenomenon. The jinx implies that whomever appears on the cover of the magazine – due to exemplary performances the previous week, month or year – will suffer equally poor performances the following week, month or year. Clearly, that doesn’t always occur.
So, regression to the mean really describes the tendency of things to go back to “normal” or to return to something close to the relevant statistical average. For instance, if you have a cold, you are outside of the average in terms of health, and the cold will eventually go away and your health will return to normal. Simply put, it’s the natural tendency of extreme scores to come back to their mean (or average) scores.
These regression effects are present everywhere. Again, sports provide great examples. In football, when a team like the Green Bay Packers, which tends to win most of its games every season, experiences a three-game losing streak, the media begin claiming that the franchise is facing a major crisis. However, after a few more games, the team most likely rights the ship and returns to its average number of victories by the end of the season.
We often experience difficulties in recognizing regression to the mean because our minds are strongly biased toward causal explanations. Humans yearn to believe that there is a causal effect behind everything we detect – or want to detect.
The harsh reality is that most of us are far too unwilling to believe that many of these effects are simply a consequence of randomness. If not, ask yourself if you would buy into a newspaper article in Green Bay with the following headline:
“Despite Three-Game Winless Streak, Packers Likely Will Come Back to Their Average Number of Victories, As They Have Simply Experienced Some Unfortunate Games”
Most fans wouldn’t appreciate such a headline, because they wouldn’t want to believe that losing could be due to pure randomness. Instead, they would prefer to find a causal explanation, such as poor coaching, lazy players and so on.
Additionally, the consequences of these effects go beyond our personal awareness. If a laundromat owner decides to implement a costly marketing program because he or she is nervous about a drop in sales – and if it works to some degree but not to the degree expected – that owner might tend to believe the expensive marketing program was solely responsible.
However, this causal explanation of the good results might likely be a serious flaw, when in fact the sales bump could simply be a consequence of regression to the mean. Hence, if the owner simply waited out the sales slump, it’s extremely likely that the results would have been the same – minus the cost of the marketing program, of course. The guiding principle: patience is the weapon of the winner.
How can you become more aware of (and comfortable with) randomness as being responsible for many life events, including your laundromat’s ups and downs? Here are two thoughts to keep in mind:
Correlation is not always causation. Despite the fact that two variables might be present in a relationship, this doesn’t necessarily mean one is causing the other. On commercial flights, I’ve often noticed that, as soon as the flight attendants serve coffee, the plane often encounters turbulence. As an over-the-top example, I might infer that serving coffee on an aircraft causes turbulence, which is a completely ridiculous conclusion. You get my point.
The more extreme the original score (sales, in your case), the more regression you should expect. If your sales were extremely poor, the better the improvement you can expect. If sales were average, normal improvement can be expected.
Try to become aware that life – including your business life – is full of luck and fortunateness. Learn to deal with it. It sounds easy to make this a part of your logical thinking, but it’s not. It takes effort and patience.
Above all, learn to accept randomness – and the fact that the Packers are likely to win on any given Sunday.
