Overall my students were pretty engaged in this lesson. However, most thought they had a working system when they actually didn’t. Common mistakes were not having a delimiter (some way of separating one letter from the next) or in some cases just repeating the same symbol for multiple letters (say using just a pig for A and F and R and X).
I thought they would discover the faults of their systems during testing (as is the purpose of testing) so I resisted just pointing out the flaws. Instead, their tests just cemented their beliefs their systems work. Apparently, students would be partially guessing as they pieced together the message, using context to correctly deduce the word.
I tried a couple approaches to resolve this…
- Tell the students their messages might need to be sent in another language or as secret code or the message could be some acronym - meaning the message may look like a sequence of random letters that don’t spell anything. I told them to try and send a message composed of random letters. In some cases, this made some realize they can’t distinguish between “pig-pig, pig” vs “pig, pig-pig” without a context. In other cases, the letters they chose coincidentally still had only one interpretation. Thus, the next approach…
- I would be their “tester” and intentionally send a message that had multiple interpretations. It bit more forced, but more certain and more efficient at the cost of a more authentic iterative process.
In either case, I would first recognize their current accomplishment of creating a system that works in many cases, then present an additional challenge that they revise their system to meet increased demands of random-letter strings. I think it’s important to recognize that they for the most part actually did solve the problem as presented (as I interpret… to send and receive actual words) - that’s an accomplishment. My students start getting resentful when I say “Oh no, that doesn’t count… you didn’t really solve the problem.” Kinda throws their work in the trash. :oP :o)
(On a side note, one group created a way to delimit characters without using fixed-length combinations - they used one animal as the delimiter animal, then used the remaining five animals in combinations for the actual letters.)