One evening at our usual beer meetup, we thought about some weird social experiments, somehow related to software development and developers life, that we want to try. Then, back from Goto;Aarhus Fulvio was talking about the conference with a psychologist friend and somehow the topic shifted to gestalt, then to the experiments we thought of. And he was enthusiast to hear about that.
So in the optic of building a cross-functional team Maria and Andrea joined us in this project. A few weeks later we arranged a meetup between developers and psychologists to mix viewpoints, experiences, knowledge, some beer and create (we hope) something new, interesting and exciting.
We wanted to find a way to start a discussion without putting some bias in it from the start. Why? One of the possible use cases is during a job interview, to understand which kind of approach a person has without forcing the conversation. We saw a similarity with the Rorschach test, and decided to try inventing an analogous test using code fragments instead of inkblots. So this is the idea with we started the meetup, and everyone got enthusiast with it.
After stating the goal we tried to learn something from our reciprocal fields, so we had a little chat about how and why we (developers) had these ideas and what they (psychologists) thought about it, taking the chance to learn and explain our viewpoints and values through examples and metaphors. When everyone got used to some basic ideas and terminology we started deepening our knowlege of the Rorschach test. Luckily Maria is currently studying exactly that test so she explained everyone how it works and which principles it's built upon. In particular we were curious about what information we could get by administer the test and which resources are needed to build it. We learned that after running the test, in order to get the results, the test subject's asnwers are given a set of scores, then checked against tables built from clinical data.
So in order to build properly this new test we need data first. But the objective of starting an unbiased conversation using code fragments is met even if it's not properly a test, hence we decided to begin using this tecnique "as is", and in the meantime gather data for a future evolution.
That was it for the meeting. We left with a target for the next time: start to build the basis for the experiment. We have still no idea about how to properly build an "item" for the test (that is, in our case, a code snippet) so the homework for the next meetup is to build some samples to administer against each other in presence of Maria and Andrea, so that they can evaluate which building principles seems effective and which not.