If you didn't know it yet, statistics is really important in science and society.
It used to be used mainly for hypothesis testing, but in these days of Data Science, Big Data and the Internet of Things it's being used just as much for making discoveries and formulating new hypotheses.
Until just a few short years ago biology was a data-poor discipline, but since The Human Genome Project was completed in 2000, biology - and bioinformatics in particular - has become extremely data-intensive and statistics is becoming ever more important.
Medicine is becoming increasingly personalised, driven by the huge amounts of data that are being captured (and statistics used to analyse them), leading to better predictions of disease risk, treatment and the quality of care.
In the wider society, statistics and data mining are being used for crime analysis, predictive policing, sentencing and crime policy. They are also being used to improve efficiency in national power grids, traffic predictions and prevention of terrorist attacks.
NASA use them. So do the United Nations, the World Health Organisation and Médecins Sans Frontières. Heck, even politicians use them to shape their campaigning strategies and guide policy.
Unfortunately, there are many Data Scientists who think that stats isn't important and we can sort of, um, do without it. You can read my thoughts on that in my blog post Statistics is Dead – Long Live Data Science…
As you've probably guessed I don't sit on the fence on this, I'm very much in favour of stats.
Despite this though, and despite the fact that I've been hanging around stats for the best part of 30 years (oops, I'm giving away my age here), I feel that - in general - stats is taught really poorly.
Over the years I've had over a dozen different teachers, and they've all made the subject feel like it's as boring as hell. Their delivery has been dire, their materials just scraped from academic stats books and they made it clear to us that they didn't want to be there and teaching was just getting in the way of their research.
Worst of all, they had no empathy. They were experts in stats, so they couldn't understand why their students were having difficulties with it - and they didn't care.
Teaching of stats is better these days though, isn't it?
Well, apparently not. I've spent days researching stats courses at Udemy to recommend to our readers and, although there are loads of courses available, I often found myself nodding off.
Not to worry though, I did find a few courses that were well presented, well taught and well thought through, so I thought I might introduce them to you.
Sitting comfortably? Then let's begin...
Delivered by 365 Careers, who currently have 20 courses in a variety of subjects (not all are data science related), this 4 hour course will help you get started with basic statistical concepts, such as descriptive statistics, distributions, confidence intervals, hypothesis testing and regression analysis.
The course is delivered with animated videos and a slick presentation style. There are lots of examples and you won't need to have had any previous stats knowledge.
In our view, Statistics for Data Science and Business Analysis is well worth taking. If you're a beginner to stats, the pace is just about right, the key concepts are taught in a visually intuitive way and it's all very easy on the eye.
This course is taught by Dr David Tan of the University of New South Wales in Australia. He may only have 3 courses right now (only one of which is data science related), but they are very well-received by his students - he has an average rating of 4.8 out of 5 for his courses.
This 4 hour course will help you get a good grounding in hypothesis testing, and takes you through concepts such as the null hypothesis, 1-tailed & 2-tailed tests, distributions, z-scores & p-values, the Central Limit Theorem and lots more.
David uses a lot of animated videos and loads of examples to explain the concepts. He even includes quizzes and practice questions with worked solutions.
The Simplest & Easiest Course on Hypothesis Testing is a well taught course and is particularly well-suited to beginners.
Kirill Eremenko is a bit of a superstar on Udemy, boasting 30 courses and being one of their top selling instructors. Most of his courses are data science related and they are highly rated by his students - and he has almost a quarter of a million of them!
This 7 hour course will teach you about distributions, the Central Limit Theorem, hypothesis testing, statistical significance, t-tests & p-values, the Pareto Principle and more.
Unlike the courses discussed above, this course is less animated, but is nevertheless well taught. He even gives you homework to do! Yay - just what we all need, right?
Statistics for Business Analytics A-Z is a well taught course and is useful for those of you who need a more business-oriented introduction to stats.
This course is taught by one of our friends Minerva Singh who is a PhD graduate of Cambridge University. Just to make us all feel even more inadequate, she also did a Master of Science degree at Oxford University. Worse still, she's a bestselling instructor at Udemy. Ah well, I console myself with the knowledge that I have achieved something of far greater significance - I have a larger following on Twitter. Thank the Gods for small mercies...
This 9 and a half hour course will teach you about applying statistics using R, and takes you through data handling in R, descriptive statistics & plotting, distributions, hypothesis testing, regressions & correlations, multivariate analysis and more, all the while illustrating how to do each of them in R.
Applied Statistical Modeling for Data Analysis in R is a must-see for those that need to learn how to apply statistics in R, although you will need to understand the basics of R to get started.
Although in general statistics is taught poorly, it doesn't have to be - and these courses prove it.
I've always thought that stats was a lot easier than it seemed in the classroom, and these teachers have found a way to make it more interesting, engaging and - dare I say it - even a little fun!
Ready to get going? Well here's a reminder of the 4 Courses Of The Apocalypse (*groan*):
There are loads of statistics courses at Udemy, not just the ones listed above. If none of these take your fancy, have a look around and I'm sure you'll find others that might just hit the spot.
If you discover any better courses out there, let me know - I may write about them in another blog post!
Final word - when you've done any of these courses, please return and leave some feedback and a review in the comments below. If you loved the course, great - come and tell us. If you hated it, that's great too - leave a comment saying what you didn't like about it.
At the time of writing, these courses are deeply discounted. They are usually offered for up to £/$/€ 200 each but are on sale right now for a few days at around £/$/€ 10 - so grab them while you can!
A quick reminder - once you've enrolled for a particular course, you get lifetime access to it, even when the course is updated.
This blog post is part of a series on learning to be a Data Science Ninja - check out the other posts in the series below:
- Learn to be a Data Science Ninja - The Easy Way
- The Top Data Science Courses at Udemy
- 4 Popular Statistics Courses for Beginner Data Scientists
- The Easy Way to Learn Python for Data Science
- How to go from Newbie to R Programming Ninja
- How To Become a Neural Networks Master in 3 Simple Steps
- The Top 3 Data Visualisation Courses at Udemy