Benjamin teaches statistics in Holland and runs his own teaching company. He loves getting students excited about statistics and math, and uses DataCamp both for himself and for his students.
How did you first get into teaching?
When I was young I taught fellow students math, and then when I was 21 I started studying and I wasn't very good at school. I was a working kind of guy, I was always working outside school and never doing my homework. But I made it through school and I started studying psychology at the university. And again, many students were asking me to explain the statistics. And then it worked so well, that I started a company. And now I've had my own teaching company for fifteen years or so. I also teach at a university here in Holland.
I just love teaching statistics or math. For most students, I see it as a kind of therapy because they are insecure about math, but when they find out that they can actually reach some new knowledge, some really difficult knowledge, then they become more happy and more secure about themselves. And I see that all the time, I see them changing in my classes, I see people coming in with sad faces and when they go out, they are smiling (because they understand something that used to be difficult). It is great to be a teacher like that. And usually they hate statistics, but they start liking it at some point. Many of my students thought they would never make it but ended up doing a Research Master's or starting a PhD.
How have you used R in your classes?
I used to do a lot of SPSS, and I didn't like it. Simple graphs are possible, but when you're dealing with more complex research questions, let's say involving four variables, you cannot graph it. Just terrible. A lot of the times, the data wasn't readable by SPSS; it was a big hassle. But then I encountered R, and quickly I fell in love with the beauty of it (also, I understood that I needed a few days to start understanding the language). When you really have to think about your analysis and not just press buttons like with SPSS, it gave me so much more insight. I think programming is very nice, but I'm a bit old fashioned. In that sense for me it is difficult sometimes to understand computers. I recently finished the Data Scientist with R track, but I am still excited to learn more!
Wanting to learn more is the best feeling. DataCamp showed me so many more options… Actually, I had the courage to say goodbye from one of my jobs because now I understand that what I've learned at DataCamp is worth a lot.
I have started teaching my students in R now, simple things. I taught a course on business analytics. We did text mining, all the basics. When we started, a lot of students hated the course, but along the way a lot of them loved it. They had so much fun, they were like, "RStudio kicks ass!" And they are going to use it now, after the class. If only 3 students out of 100 continue, I'm already so happy.
I'm really not a computer person, but DataCamp made R accessible, intuitive, and fast for me.
How do you convince skeptical students?
Well, it's usually by asking interesting research questions, which they usually think that they have an answer to, and then showing them some good graph and they see that the whole world is actually completely different. Most students don't understand interaction effects. And I think that is one of the most complicated things, to understand an interaction effect and really feel it. And when I tell them about these effects, then they start thinking differently and they realize, ok, we need something to actually test that. And what are your options? You can do it with SPSS or you can do it with R, but eventually with R, if you use ggplot you get such nice graphs. You can't do that with SPSS. And they are soon convinced. But of course, a lot of students have problems, and they have fear because it is programming, it is abstract. But that is also what I like, to convince them. I don't want easy students.
How did you come across DataCamp?
Just online. I tell everyone nowadays about DataCamp. I just say, "Go on DataCamp, you'll find your way." Many people I know are also addicted to your platform.
When I have the time, I would love to use the Teach app to make a Dutch course with DataCamp for stats-oriented psychology students. That is my biggest wish.
Do you think a DataCamp subscription is worth it?
So I started with a course on Udemy. I was really irritated because I was being spammed, and I thought, this is not what I want. You shouldn't force me to give a critique, so I left immediately. Then I found DataCamp and the whole environment worked so well, that I immediately paid a full year contract. I thought, "This is a good environment to learn." I was frustrated with psychology or statistics in Holland, because it is a lot of theoretical based modeling, and they do a lot with numerical data and they are skeptical about text data. But when I start thinking about the world around us, I think there is nothing much numerical about it, it is more or less discrete and it's nominal in a way—text data is everywhere. So I want to see the world in a nominal way, and not only in a numerical way. And I think R will help me with doing that.
What do you like about DataCamp?
The setup: on the left side you see the question, on the right side you can type your answer, and the console, and the slides, and the suggestions you get when you do something wrong. With all that together, you don't need to go somewhere else to figure it out. So it is completely self contained, and that is a very good thing. The movies you make are beautiful, you have nice visuals. It is easy going, no weird, wild stuff—it is just how I want to be taught.
I had the courage to say goodbye from one of my jobs because now I understand that what I've learned at DataCamp is worth a lot.
Have you been able to use any of the new skills you've learned?
Oh yeah! Actually, I recently used Ted Kwartler's text mining course. I used most of his stuff for the business analytics course, the codes from him and stuff. We had reviews of all the restaurants in Leiden—the city where I teach—we had 12,000 reviews, and we had some metadata on the restaurants. So we did text mining on the reviews with some business analytics. So for the students, I had a subject which was interesting. It was their restaurants, their city, their stuff. So they were eager to find out stuff about the restaurants they normally visit. So in that sense, it was a real life thing. So yes, I do implement them all the time.
What was the hardest thing for you when you first started?
Programming. Definitely. Working with dplyr and stuff, that is ok. But the for loop, the while loop, that was sometimes hard to get my thoughts around it. But with practice, it was fine and I was able to figure it out. But the programming aspect was the biggest challenge for me. I am mainly interested in data: I want to analyze and learn about data. And now I realize, programming is necessary to do that, if you want to do it efficiently.
How has DataCamp helped you learn programming?
Mostly it's the environment, which is so well designed. The fact that you get helpful feedback. It's very frustrating for a beginner to search online when you don't know where to search. Of course, searching yourself will suffice for those who are disciplined or who have plenty of time. But my life is pretty jumpy and I'm running around most of the time to do my work (which I still love to do).
Also, the fact that after each exercise (which take at most five minutes each), you'll get that nice Sudoku feeling—everyone loves it when they solve a problem. In that sense, doing courses at DataCamp is also pretty addictive, which is good in this case. The level of difficulty increases with steps that are not too big or too small. And if there is a big jump in difficulty, you get some clear suggestions of which courses to do first.
Your instructors are inspiring people, authentic and nice to listen to. Maybe it's because I'm biased by now, but I never got irritated with any of the instructors at all. Some of them are actually pretty funny as well!
When I encountered R, I fell in love with the beauty of it.
How do you feel DataCamp balances theory and practice?
It is good to have some theory. You did this really well in the statistics course with Daniel Kaplan, where he had the "Null Doctor". That is amazing! That is nice for educational stuff, for people who aren't directly into stats. So in that sense, DataCamp is great for both the people who already are familiar with the theory and need the practice, and it is great for the people who are new and it helps get them excited about data science. Of course you want to learn things fast, but you also need to grasp the theory if you really want to understand something.
It sounds like DataCamp has helped you as a learner. Has DataCamp also helped your teaching?
I use my knowledge from DataCamp all the time. Sometimes, I will use the explanation of a concept from a video in my class. I already knew statistics, but I share the programming and the data manipulation techniques that I've learned on DataCamp all the time.
I've also learned to be more free in my thinking because of DataCamp. Theoretical modeling in psychology and machine learning have such a different approach. I am starting to use more machine learning in my work more nowadays. I tell my colleagues, "Well, no, we're not going use your theoretical model, we are going to try some different models as well, and see what the data tells us." That is a perfect example of the type of things I regularly use that I learned on DataCamp.