Simply Explained Data Science Concepts:

Describing Technical Concepts To My Children…Part 1

“If you can’t explain it simply, you don’t understand it well enough.” — Albert Einstein

Learning new information can be an arduous and time consuming task. However, the ability to transfer knowledge is what separates us from other living beings. It is a beautiful process that has given rise to our technological advance society we currently live in today. Furthermore, learning new information is one of the cornerstones of our society. Thus, in order to truly ensure we have learned said information we should be able to teach another person the aforementioned information. Therefore, I have taken it upon myself to teach my two children Data Science concepts in terms they can relate to.

With that being said, my children are aged 12 and 3 respectively. Now you may be asking yourself, “Is that a typo, did he really mean to write 3. A 12 year old is understandable, but a 3 year old!?!” and you’re absolutely right to undermine my ability to teach these concepts to a 3 year old, but it won’t stop me from trying. Furthermore, while it is abundantly clear my 3 year old won’t be learning the fundamental concepts of Data Science, I will still provide a learning experience in the form of listening, answering questions, and providing undivided attention. As for my 12 year old, the teaching/learning duality will be in full effect.

Topic 1: Hypothesis Testing & T-tests

For hypothesis testing, it is imperative that you first start with a question about a large group. For example, what is the average weight of U.S. citizens? A typical guess would be 180 pounds. However, you know that you can’t realistically measure the weight of every single U.S. citizen, so you randomly pick a smaller group to measure, a random sample.

To see if the average weight of all U.S. citizens is really 180 pounds, you would conduct a hypothesis test! In this example, you would need to use a one-sample t-test. The important thing to be aware of here is that you need to know the number of people you weighed, their average weight, and the standard deviation of their weights, how far away their weights are from the average weight.

When you get a value, t-value, from the t-test, you need to use a chart or program to evaluate whether or not you can conclude you are pretty sure — 95% sure — that this is the average weight of U.S. citizens. If your t-value is larger than 1.96 or less than -1.96, you would say that you reject your hypothesis that the average weight of all U.S. citizens is 180 pounds. If it is in between that range, you “fail to reject” your hypothesis and conclude this is the likely the actual average weight of U.S. citizens.