The key difference between a data warehouse and a data lake is how the database is structured and how it's used.
Think about the last time you heard someone admit to being wrong about something important and personal — not wrong about something insignificant, such as the scheduled time for an event, but rather wrong about a passionate belief or an ingrained personality trait. Can you think of any? If you can’t, that’s okay — these occurrences are rare. A person's identity is strongly associated with his or her personality and beliefs. Admitting to a fault or flaw in these areas is a challenge to the person's identity — not quite the person's existence but certainly his or her essence.
In an organization, leaders are often unwilling to admit their errors or uncertainties, because they are afraid that by doing so they will be perceived as weak or indecisive. If you follow politics, you can witness this phenomenon on both sides of the aisle — politicians who make decisions and embrace certain positions not because they believe it is best but because they are afraid of being perceived as weak or uncertain. They even go so far as to spin the facts to support their respective positions.
A University of California physicist named Richard Muller spent years arguing against global climate change. He helped found the group Berkeley Earth. Much of his work was funded by the gas and oil industry. Later, his own research found very strong evidence of global temperature increases associated with human activity. He concluded that he was wrong. Humans were to blame for climate change. Muller saw that the facts against is belief were too strong to ignore, so he changed his mind. He didn’t do it in a quiet way. He wrote a long op-ed piece in the New York Times that outlined his arguments and why the counter arguments were stronger.
The most effective leaders are actually those who are strong enough to admit when they are wrong. They are rational and make decisions based on information rather than opinion. They do not get defensive when challenged with facts that counter their assumptions.
Challenging someone else's beliefs or assumptions is relatively easy compared to challenging one's own. The distinction can be attributed to two types of critical thinking:
There is no shortage of people who primarily engage in weak-sense critical thinking. They are the people who strongly defend their own positions and equally strongly attack the opposition. They're not very good at challenging their own positions or recognizing any merit in opposing positions. When losing an argument, they get defensive and emotional and often irrational because they so closely identify with the position they hold.
Strong-sense critical thinkers are rare. They are the people who, when confronted with information or opinions that contradict their beliefs or assumptions, are willing to listen to and explore other possibilities. They ask themselves, "Could I be wrong about this?" and "What if I am wrong about this?" They look at the facts, question their own assumptions, and pick apart the logic of their own reasoning. They are committed to the truth. These are the people you want on your data science team.
Imagine how different levels of data critical thinking might play out on a data science team. Suppose a running shoe website runs a promotion and sends out a coupon to everyone who buys a product. The data science team looks at the number of people who used the coupon to make a purchase, and the team produces the data visualizations shown below.
The graph on the left shows that more than half the customers received coupons, only about eight percent clicked on the coupon, and only about half of those people used the coupon.
The graph on the right compares coupon and no-coupon sales. Notice a few spikes in coupon sales primarily the day the coupon was issued and a few days afterward. Also notice that the coupon seems to have increased both coupon and non-coupon sales, but coupon sales account for a relatively small percentage (about 10 percent) of the total sales.
Your data science team wants to determine how successful the coupon was in increasing revenue. Of course, the team could simply look at total revenue in the 14 days prior to the coupon release and revenue in the 14 days after its release and compare the two numbers. However, that would shed light only on whether the coupon was effective and to what degree. It wouldn't explain why the coupon was effective or whether other, less costly, promotions would have been just as effective if not more so.
This is where strong-sense critical thinking comes into play. The data science team should be willing and able to ask more probing questions, such as the following:
When your team applies strong-sense critical thinking, it should feel more like an open discussion. No one should feel as though they’re defending themselves. This approach is a great way for your team to ask interesting questions and in the end, gain greater insights.
The key difference between a data warehouse and a data lake is how the database is structured and how it's used.
A database schema is the organization of data in a database. They are usually found in a relational database management system.
Big data applications will only give you useful insights if you have access to high quality data. If you're finding a lot of data garbage then it will make it difficult to learn something new.