The Art of Learning Statistics and Resources Required
A comprehensive resource guide to learning statistics effectively and thinking like a statistician
What is statistics and why learning statistics is a good idea for data science?
In simple terms, statistics is the art of analysis and interpretation of data using mathematical tools. For example, During Covid, we have seen that multiple vaccines have been introduced claiming x% of effectiveness against Covid with 99% accuracy. Are you wondering how pharmaceutical companies estimate these numbers from clinical trial data? Simply with the help of statistics. Isn’t it a good motivation to learn statistics so that we can estimate important key aspects from a bunch of data?
Important topics, one should learn to start their journey with data science
I have tried to group essential topics of statistics that will be enough for anyone who is starting his journey in data science. Statistics is a vast domain in itself, so I have covered essential topics that will give you a good base and confidence to deal with day to day work of data scientists. In simple words, I have picked the river of statistics that are required for survival but there is a whole lot of ocean left which we can learn going forward on a need-to-need basis. All mentioned topics are covered in the given resources.
- Introduction: Key definitions, types of Statistics, sampling techniques, etc.
- Central tendency metrics (Mean, Median, Mode)
- Standard deviation, Z-scores, percentiles
- Basic understanding of probability
- Correlation, analysis of variance
- Probability Distribution: Binomial distribution, Normal distribution, Random variables
- Sampling distributions, central limit theorem
- Finding Confidence intervals with sample and population statistics for proportion, mean, variance & standard deviation, Usage of Z-score, and T-score.
- Hypothesis testing for mean, variance standard deviation.
Best resources to learn statistics effectively
1. Professor Leonard (primary): If you are someone who likes the classroom lectures, then I would say it will be a perfect resource for you to learn statistics. The professor is very good and teaches a subject in a very interesting and simple manner so that you grasp the topics quickly.
2. Online courses (primary): If you like to learn from online courses then the below two courses from Udacity are great to learn statistics. Start with the descriptive statistics course first (builds the base for the second course) and then take the inferential statistics course. You can audit the course for free access to course resources.
a. Descriptive Statistics course
b. Inferential Statistics course
3. StatQuest (secondary): Joshua Starmer is a well-known YouTuber in the field of statistics. He teaches even the complicated statistics concept in a very effective and visualized manner. Visualizations help in understanding how that concept is built piece by piece which ultimately builds your intuition about that concept. Intuition building is key to learning as formulas can be forgotten but intuition remains standstill.
“The only real valuable thing is intuition.” — Albert Einstein
4. ZedStatistics (secondary): It is another great YouTube channel that has some great content on statistics. This channel posts videos that cover specific statistics topics in detail from visualization of concepts to the history of topics and discuss even advanced concepts. If you do interested in advanced stuff for given topics, do check out this channel, especially the chi-square distribution video is a gem.
To effectively use the above resources, I would recommend picking either Professor Leonard Course or Udacity Course to learn statistics in a structured way as they both have a well-laid out syllabus that helps in tracking progress. While learning from either of the resources, probably you may have some doubts about understanding some concepts or visualization. To cover those doubts, StatQuest and ZedStatistics channels come in handy as they have detailed videos on almost all topics. In this way, we’ll have two levels of learning resources (primary/secondary) that’ll save a hopping time from one resource to another in case of doubts.
Two are better than one… for if they fall one will lift the other.
Let’s go ahead and embrace your statistics learning journey!
