In all my years teaching Statistics, there is one thing that has been and will constant: students will always get lost. No matter how I teach the subject, what kind of textbook I use, what kind of approach I use, it is always the same. Kids get lost and they will end up needing help with their statistics homework. And the trajectory of such phenomenon is also very predictable: First they start off really well. They are happy because they find the topic of descriptive statistics very easy. Then, when it is time to go onto normal distributions, they start to feel the heat bit.
It is frustrating for me as an instructor. They will sit with me during office hours and I will go through their stats assignments and I will give them pointers, and that is enough. They learn normal distributions. But when it is time to do hypothesis testing, then they go crazy, they get lost especially with the two populations (understandably so, because such tests tend to be algebraically more complicated than the one populations test).
They Know How to Plug Values in But they Don’t Understand Concepts
One of the most frustrating thing for me is to see how students get lost and simply do not understand the idea of null hypothesis. Overall, they have difficulties to understand why I assume the null hypothesis (Ho) as a premise, in order to compute a p-value. The p-value is probably the most misunderstood measure in statistics, and it gets lots of students confused. Help with statistics homework anyone? I thought so.
I have heard everything when it comes down to give wrong interpretations to the p-value. So many forms to get it wrong. One very repeated wrong answer tells you that students believe that the p-value is the probability that the null hypothesis is true……Brrrr. I have failed as a stats instructor. But wait, it is not just me. Indeed, I talk frequently to other stats instructors (and stats tutors as well), and they get that too.
Then, When The Harder Topics Comes, Anxiety Kicks In
The first part of a stats 101 class usually consists of descriptive statistics, normal distribution, and hypothesis testing. Those topics are the easiest, or at least are the one that students get a better grasp off. Truly, the other concepts involved, that come after the basic ones, such as ANOVA and Regression Analysis, are equally simple, but kids struggle more, no doubt. And then is where they will need more help with their statistics problems. Not because those problems are harder per se, but because those are less commonly known.
My advice to my students is always the same: resilience is the key. What really pays off in the end is to continue the learning process, even if you get lost, and fall behind. You will need to learn to work under pressure, and believe me, that will pay off. Ultimately, you will acquire some proficiency, maybe not enough to get an A, but you’re guaranteed a passing grade.