“You know what, the Quadrilaterals activity never really convinced me,” a colleague of mine told me when I recounted to him my experience with the activity at the primary school in Radenci.

But let’s start from the beginning: two school lessons, two completely different stories.

We first experimented with the Quadrilaterals in the seventh grade. The students were divided into pairs or groups of three, and each pair or group got its own tablet, which they used to enter the characteristics of quadrilaterals, and classified them into squares, rhomboids and so on. No different from what we typically do in this activity. Except that it somehow… didn’t work. The students were reserved, they didn’t reply to my questions, didn’t ask anything… If something like that happens to me in a classroom somewhere in a region like Prlekija or Slovenske gorice (located far away from Ljubljana by Slovene standards), my first thought is that perhaps my “lublanščina” (the infamous Ljubljana slang) is to blame. I then switch to some kind of semi-formal Slovene that probably sounds even weirder, because nobody speaks like this.

The fact that two student groups accidentally registered into the data entry website twice – in this way unintentionally taking the spot of other groups, of course didn’t really help either. The software is supposed to prevent those kinds of complications, but this time it didn’t. (The issue has now been fixed. “We apologise for the inconvenience and kindly ask for your understanding,” as they would say at the Slovene national railway company after each late ride - which is, more or less, always.)

After that experience, we started the next lesson that followed right after in the sixth grade, with a bit of a lump in our throats. The topic of the activity we were about to do was Animal Classification Keys. While we had already tested a similar activity called Zoo with students several times, the Animal Classification Keys had only been tried in class once and only briefly. And yet! It worked like a charm!

The first thing that we did differently was: the introduction. I made up some story about an evil dictator somewhere in Finland that holds you captive and will chop off your head if you don’t manage to correctly classify 13 animals into animal groups like mammals, insects, and so on, by the morning. The catch? All these animal names, characteristics and categories are given to you in… Finnish! (A language you probably don’t speak masterfully . At least not many Slovenes typically do!) After the initial objections by the students and making it clear that using the dictionary or Google was not an option, we got down to work. The students sat behind the computers, and all of us were using Orange to solve the challenge. With the Distributions widget, we looked for the animal characterstics, based on which animal types (in Slovene) differ from each other; we built a classification tree on the whiteboard, and then let the the computer continue with the tree building process. We used the Prediction widget to make sure the tree worked corrrectly, and then, finally(!) we were brave enough to attempt to perform the task in Finnish.

I know, this description might seem a little short and perhaps it’s not perfectly understandable. Never mind, we will describe the whole process again in detail in the adapted description of the activity, where it will be much easier to follow step-by-step. But what is more exciting is how engaged the students were, how well they understood classification trees and the algorithm that builds them, and even the principles of working with Orange! It all made sense to them, and everything came naturally! When, for instance, the Predictions widget wasn’t showing the results at one point, one of the students suggested plugging the tree building algorithm into the Prediction widget’s input. That is something that not even many university students would understand when first encountering Orange, and many don’t even grasp that connection later on, or ever!

“You know, the Quadrilaterals activity never really convinced me,” a colleague of mine said. Well, I still think it is fun: the topic has a very clear place in the curriculum, the activity is suitable for revising the shapes, and so on. But it’s true that every now and then, an activity flops in the classroom. What the Radenci experience made us realise is the importance of an attractive story that can serve as a motivation for the students to engage with an activity. Like the one we used for framing the Animal Classification Keys in our second experiment in Radenci. That element is still missing in some of our activites.

To the teachers in Radenci, Jana Grosman and Simon Belec, thank you for your warm welcome!