Statistics is one of the most powerful tools for driving change in any Business, however, if misused or misinterpreted, it will lead to producing biased results.
Let’s split the misinterpretation of statistics into 3 main points for better understanding, with real-life examples:
- Bias, non-random and small samples:
Having 3 out of 4 rich adolescents from New York city telling that the iPhone is their preferred smartphone can never be used to tell that 75% of people prefer iPhones over other smartphones; the sample is too small, bias and not random.
Asking 4 out of 7.5 billion people, who live in the same city, have the same age, and belong to the same social class can never provide useful insights.
- Wrong chosen average:
Having an average monthly salary of 100 employees in an IT firm equals to 5000$ won’t tell that you will have 5000$ a month working in this firm. The Mean (arithmetic average) is useless here, when 90 employees have monthly salaries between 1200$ and 2000$, while the directors and executives have monthly salaries between 20000$ and 30000$. The Median (a middle point with half salaries below and another half above) and the Mode (the most common salary given) in this case will give more realistic information and a better idea about what your salary will be if you will be employed.
- Semi-attached figure:
After giving to try and asking a very good number of locals about the food in a specific restaurant in Moscow, they all answered it was very healthy, very delicious, and not expensive. However, this same restaurant is not able to attract more than a couple of clients per day. Omitting the fact of the not so good presentation and the missing soup in the menu which is essential for Russians at lunch impacted all the analysis done.
Computer Engineer • Entrepreneur • Blogger