It’s one thing to read about a statistical model or methodology. It’s another thing to explain it to someone else. My theory on learning is try to teach it. There is something about saying it out loud that awakens the brain. You actually hear yourself navigating through the logic in your text-book. A semi-formal or organized presentation helps you to think clearly about a statistics topic. What is the purpose of a statistic parameter? What characteristics define your test statistic? Can you run through an example? Is it possible to apply your logic to a completely different scenario and then check your results?
A couple of years ago, I took a proof writing course from hell. I truly mean it when I say it wasn’t from heaven. The text-book was of no assistance. But it helped to talk out the logic on a white board with an audience (preferably peers). Your listeners can provide you with feedback. Note: I strongly suggest you ensure there are some quasi-experts in the audience (peers who know the subject better than you). I am just keeping it real here… The last thing you need is an audience that “co-signs” – on “facts” which are not accurate.
Reading provides you with enough background to ‘speak’ on a subject. But in statistics, you must be able to synthesize concepts in a way that makes sense. So it’s not enough to know an equation exists on page 5. You should be able to break down what every symbol represents; how it is used; and provide examples.
Now I know some of you are reading this like – why do I need to do this?! You don’t need to take my advice. I am just suggesting that if you have trouble studying – it might be helpful if you form a study group and talk these things out!
If you enjoyed this blog there are more to follow!