## Outliers: Definition and Sample

Outliers. In statistical terms, outliers is an event that deviates far from the mean, where most events or things are expected to happen or be in. Sixty eight point two percent of events lies between the mean and one standard deviation, this is the general population as we know it, they are people who define what is normal and what is not in the society we live in. Ninety five point four percent events are expected to be in between mean and two standard deviation, in this broader range, it includes semi-normal people in population, they may be a little weird and unusual, however the society still accept them as one of their own. Then there are outliers, the rest of the population not included in the category of “normal”, the way society defines it, the rest four point six percent.

The graph above is called the normal distribution probability, this graph represents almost everything we know in life regarding a chances. Let’s look at several examples, the first one taken from the average household income for US in 2010. The average annual income for every household in US is \$67.530, meaning that if we are guessing our friends or family’s annual income, there are big chance that their income will approximately be around that number. From the graph below, we could also say that there are 95% chance that people you met today has an annual household income between \$10000-\$200000.

There are, of course, millionaires and billionaires whose income are above \$1 million per year, these people are outliers because they deviate very far from what is considered as normal (the mean). Outliers in this case are not only the very rich people, but also beggars and homeless people whose annual income below \$1000. These two category of people, the ultra-rich and the poor skewed the distribution of the population, depending on the size of the two category. I hope by now you’ve understand the concept of distribution and outliers.

Another example I like is about people personality, in the book “Quiet”, Susan Cain mentioned that introverts are roughly about a third of the whole population. I would like to use the same concept to categorise people personality. Each of us know, in our daily life, people who can’t stop talking and people who barely talk at all. Luckily most people we met are not either of those two, in that case we have to choose between a “broken radio” or an “awkward silence”. People who have a good social quotient are capable of balancing the extrovertedness and introvertedness inside them, they know when to apply both characteristic in various situation. On the other hand, the outliers are not very good at adapting to social situation. They tend to be very talkative and enthusiastic when interacting with peers, or barely talk if not at all.

You see, most events in the world are distributed evenly and you just have to observe to know where you stand. The number of rainy days in a month, the scores you’ll got on the exam, all could be predicted using statistic and historical data.

A global macro analyst with over four years experience in the financial market, the author began his career as an equity analyst before transitioning to macro research focusing on Emerging Markets at a well-known independent research firm. He read voraciously, spending most of his free time following The Economist magazine and reading topics on finance and self-improvement. When off duty, he works part-time for Getty Images, taking pictures from all over the globe. To date, he has over 1200 pictures over 35 countries being sold through the company.
This entry was posted in Review and Ideas and tagged , , , , , , , , . Bookmark the permalink.

### 3 Responses to Outliers: Definition and Sample

1. Jim Kane says:

A great post. I have read Gladwell’s book of the same name and Susan Cain’s Quiet. Regards, Jim

2. No-hoper says:

That was a really good read 🙂 Something that these graphs don’t necessarily take into account is partial dependants; for example, the low earning outliers may be part time workers supported by a high earning
spouse, and this could distort the graph

• agent909 says:

Thanks. I agree that there are many flaws in the graph itself, that is why the statistician use household income rather than personal income.