We never ever applied the data to a practical real-world problem. We never understood what sort of conclusions could be made, and what couldn’t. We were just given random data sets and told to compute variances and means… and I was like, why bother?

Trust me. I went through Stats Honours. I came away with a lot of formulas, but I still don’t know when to use which statistical tool. My friends who use stats on the job tell me that you only learn by doing statistical analysis with a mentor who’s really good at both stats & real world analysis.

In essence, I wish my college instructors had focused on the whole learning process. Explain the objective (central tendency). Give examples of real world situations where that objective is useful (average weight to fight obesity). Go into maths of tool (mean, median etc.). Compare and contrast strengths and weaknesses of tool with respect to that real world example. Case study – full analysis homework with conclusions & interpretations also assessed. Now that would have been a great stats course.

]]>When I reached graduate school I took a couple of classes that was taught by a different instructor and it was as though cold water had been splashed over me! The new instructor I had taught the subject in such a different manner that it made things easy to comprehend and follow.

I quickly realised that the problem my first instructor had was that she focused way too much on the theoretical side of statistics (in a monotonous manner) and failed to teach using more pragmatic methods.

For example, instead of focusing on explaning multiple regression in terms of equations and the relationships between X, Y Z and nothing else and then bring up some stats output and proceed to explain it without showing us how the output came about. We would then attend fortnightly tutorials AFTER the lectures and try to run the output ourselves.

On the other hand, the postgrad instructor I had would explain it quickly in a theoretical manner, then reiterate using a real life example and THEN best of all fire up a stats program such as SPSS on the spot and show us how to crunch the numbers, generate the output and interpret it. The best part is when you can relate the output/outcome back to the theory to see how it all falls into place!

Because statistics follows logic – if one part is poorly taught earlier on, I tend to find that many students (such as myself) will have difficulty understanding many concepts later on.

]]>To most people, I’m an outlier just because I think this way. I might prove problematic to someone who needs to do some serious manipulation. In that case, I would simply be done away with. All one would have to do is to exclude me from the data. You see what I’m getting at? It can be dehumanizing. Some people get left out of the sampling process. Then their needs go unmet in favor of the so called average. It could happen to anyone, really. Read Jung’s work, The Undiscovered Self, (if you haven’t already) and you’ll know what I’m getting at, and why I personally am so wary of statistics.

I know that statistics can be useful, but the problem is that it’s such a handy tool that like some new invention, it can be put to good use, or it can cause such widespread destruction. What’s at stake here is the human soul.

What makes it worse is the entire educational system having to rely so much on the directive approach to educating, rather than non-directive, as though individuals can’t think for themselves, but need to be controlled and manipulated and kept from straying like some seemingly insignificant outlier.

Statistics is difficult not because people can’t think abstractly or can’t do algebra, but because of what it implies and the way it gets administered, and the way it takes your soul away, and the way it forces you to play the game. Nevertheless, I am learning it because I have to. I might live or I might die in the process. War is hell.

]]>Finally, i think that statistics as an academic domain is confused with statistics as a tool of persuasion (in the sense of “lies, damned lies, and statistics”). I think this has given the layman the impression that statistics is not so much a rigorous branch of mathematics, but rather a means of selecting the best data to support your own view.

]]>I hate they way statistics is taught. It is too broad of a topic for a specific course, and should rather be integrated (no pun intended) with other courses. For example, while learning integration in a calculus course, there could be a small aside involving the normal distribution (since the normal dist. integrated from -inf to inf equals 1). One of the few parts of statistics that I don’t see able to be merged into another course is probability, which could easily be fit into a ~3 month course.

I also dislike the inconsistent notation used with statistics. It’s like everything BUT statistics uses a nice, standard notation.

One last complaint; I absolutely HATE the reliance on numbers (this ties into the way it is taught). There is so much that could be done symbolically, but is instead done using brute-force methods such as using random numbers.

Hopefully I can avoid any more statistics courses until I’m done with my schooling ðŸ˜‰

]]>