Statistics for Data Science and Business Analysis | by 365 Data science on Udemy | course review

Statistics for Data Science and Business Analysis by 365 Data science

Video Review

This course is a statistics course. And yes, I think it’s the best first course you could be taking if you start learning data science from scratch.

What is this course about?

This course is a pure statistics course.  When you’ll take it, you’ll be learning:

  • The basics of statistics (So, descriptive & Inferential statistics) 
  • You’ll be learning how to make sure your observations are accurate with hypothesis testings
  • And finally, you’ll learn how to predict future possible observations and outcomes with linear regression.

Don’t worry if you don’t know what hypothesis testing, linear regression is, or anything about statistics really, because neither did I when I was first learning about it! 

What is descriptive and inferential statistics and why do you want to learn that to work in data science?

What it is? Well to put it simply, descriptive statistics uses data to describe the whole possible set of observations in a given niche, in other terms population. 

So, if you have the salary of the 10 employees of a small digital marketing agency put in a table, “That” could very well be treated as descriptive statistics because it covers the entire population of employees in the company  

But what you mostly work with isn’t descriptive statistics, what you mostly work with is inferential statistics;  And that’s especially true if you’re a data scientist and not a data analyst.

Inferential statistics uses samples to try to mirror the whole set of possible observations in a given population. If you’re a data scientist and not a data analyst what you mostly work with is inferential statistics.  

Inferential statistics uses samples to reflect the whole set of possible observations in a given population. Mastering it is obviously essential and necessary if you aspire to become a data scientist or a data analyst since it’s at the basis of any data analysis

Make sure your observations are accurate with hypothesis testing

Once you’ve learned all about descriptive and inferential statistics, you’ll want to know how to do hypothesis testing to make sure that what you found does mirror the population you’re studying . 

Let me show you how hypothesis testing works with this short example:

So, let’s pretend that you’re a data scientist and that your client tested different ads with 1000$ budget on each ad and wants to know which one works best. When he’ll know for sure which one is the most profitable, he’ll invest 30 000$ on it.

It’s with hypothesis testing that you’ll make sure the observations represent the population.  (The ”population” would be all potential clients in this example)

It’s not because one ad seems to be performing better than the others that it actually does.  Big companies would be losing millions or even billions if they assumed such a thing. 

I’ve explained it to you that way so you can kind of understand what it is, but I’m giving you a bad example because in reality you could do hypothesis testing with every observations you make. 

From good to great. Learn about regressions.

What I’ve explained to you so far really is the biggest portion of the essential statistical knowledge a data analyst must have.  The portion of the course which is going to be useful for future data scientists is linear regression

Unlike data analysts, data scientists use data to predict future outcomes. And linear regression is all about that. 

Linear regressions are easy to understand, but way harder to master. So, after taking this course you will perfectly understand how to use linear regressions. But believe me, you’ll suck at it.  What you mostly need, in order to master this, is practice.

You will have to fall flat on your face many times! That’s really how you’ll get good at it.

What this course does especially well!

Deep comprehension

It is amazing at making you understand the math and the ideas behind every shown statistical concepts. (Which is something that is lacking in many courses)

Probably the greatest first statistics course

See it really starts from the basics of statistics so even if you don’t know anything about statistics, by the end not only will you be able to think rationally and statistically as a data scientist would, but you’ll also be able to perform quite complex statistical analysis and have all the statistical tools to start working as a data analyst! (You’ll still have a lot to learn, but you’ll be able to get started)

Provide great exercises

Then, the third thing I especially liked about this course is the exercises it provided. Relevant and useful exercises that will help you understand, memorize everything you’ll be learning and that will also give you great confidence in your ability to perform statistical analysis and to treat data effectively. 

When reviewing a course, one of the things I ask myself is : how good is this course at getting someone to his or her goal? And courses use different methods to make sure they help you achieve your goals.

But, when the information is accurate, complete and well presented, what makes the difference between a good and a great course is almost always the quality of the exercises, And it was a real pleasure to witness how great the exercises were in this course.

There could have been more but who will complain? It’s a 20$ course for god sake.

Plus if you want to have more statistics exercises, there are great resources out there.

What isn’t so great about this course

The thing you might struggled the most with in this course is the mathematical part of it. The problem is in this course the instructor does not explain that much math. 

It’s really basic math, but if things like sigma are forgotten memories for you. You’ll struggle at the beginning of the course.

You’ll have to refresh your memory on YouTube a couple of times at the beginning, but you’ll end up doing just fine right after you remember some of the math principles you might have forgotten or just after you learn it if you never knew it.

The BIG issue with this course

Then the biggest problem you might have with this course is the fake monotone and a bit robotic voice explaining everything

I recommend you first listen to some of the demo videos available on the course’s page before buying.  If you get frustrated with it just don’t buy it.  Check some of the other reviews I’ve published to find a statistics course that won’t bore you. 

For me though, it hasn’t been an issue. I thought it was alright. It had the benefit of being really clear and concise 

Which leads me to another little critic which is maybe more of a warning then anything.

It goes really fast

Don’t skip any videos or any exercises and study everything before getting to the next step.

The course is built in a way that once you see something, the instructor assumes you know it and understand it before you go to the next video.  So feel free to rewatch everything as many times as you need. 

Final thoughts

Overall it’s a great course, but obviously it’s not for everyone, so here is who I’m recommending it too: Aspiring data analyst. 

Man, I mean… this course provides you with all the statistical knowledge you need to get started and it only cost 20$!

Now, an aspiring data scientist will get some tools… Like a screwdriver and a saw maybe, but you need the entire toolbox to become a data scientist. 

So yes, you learn linear regression which is one tool that predicts future outcomes, but you’ll still have a shit ton of stuff to learn before you can call yourself a data scientist. It requires a lot more time and effort to acquire such statistical knowledge. 

But hey! It’s still is a great first step if you don’t have prior statistics training.

Bonus exercises

If you need more exercises to learn statistics, I’m recommending you take a look at ”statistics student workbook” by Robert S. Witte.

It’s all just statistics exercises 😉 .

Leave a Reply

Leave a Reply

Your email address will not be published. Required fields are marked *