ScienceAtHome rewinds–Part 2
Do you remember the first adventures that gave birth to Science at Home? Today, Jacob is taking over for this post so you can get an extra insight into the ups and downs of the project’s first years. Let’s pick up where we left off last time!
The first seed money
For a while, it was just a crazy idea, but in the end of 2011, the Science faculty gave us a little bit of seed money allowing us to hire one of Klaus Mølmer’s recently graduated master’s students, Sidse Damgaard, for a month. She took over our theoretical modeling code in a mathematical program called Matlab, and constructed the first simple user interface to the physics code. With this, the first iteration of what is now Quantum Moves was born!
As Sidse’s contract ran out, we were again lucky to be awarded a bigger chunk of seed money from the Aarhus University Research foundation. This was used to establish CODER -the interdisciplinary AU Ideas Center for Community Driven Research- bringing together scientists from physics, computer science and educational science. Sidse kept expanding the interface, and from February 2012 onwards we started running the first tests of implementing the game in an educational context, in several Danish high schools. Although the interface was very basic and the didactics not at all developed, we got very positive initial responses.
Our first miserable failure
One factor, however, remained a source of great annoyance. Using Matlab got very cumbersome and it complicated our process. In June 2012, we decided to run a large beta test and through radio appearances and advertisement, we managed to recruit 100 volunteers! We were thoroughly frustrated with the immense list of tasks that had to be checked off, so we decided to go for a flexible interface. We thought it would allow users to help us decide on several aspects of the game, such as the order of the various game levels. Moreover, they could rate the importance of each level. We also wanted them to be able to design new levels in a creation environment.
The result? We failed miserably. The vast majority of our volunteers did not succeed in installing the program. And the remaining dedicated few quickly dropped out: they had no idea where to start in what eventually appeared to be a complicated user interface. Tutorials and informational material were shamefully lacking.
A fresh start
The frustrating experience led to a drastic -and wise- change: we dumped Matlab. In a heroic effort, Sidse, Tilo (a newly hired student programmer) and Lenard Sobirey (our first CODER visiting fellow from Hamburg) managed to translate the entire code into Java, a more flexible programming language. And they achieved this task within just a couple of weeks, during the summer of 2012. Since then the framework has grown step by step into the current Quantum Moves.
The path back then was not easy. Being a team of happy amateurs, we had never developed a game before. Our seed funding was intendedly interdisciplinary and the underlying aim was to attempt a systematic design project through which we could share our experiences with the rest of the citizen science community. By doing so, we could then offer small assistance to future citizen science programmers out there.
In this quest, we were immensely lucky to get Andreas Lieberoth -an educational psychologist- to join our team shortly after. Andreas, unlike the rest of us, had large experience in game design. Furthermore, he was extremely dedicated to systematically testing explicit and implicit design hypotheses and assumptions. Thanks to his help and in a collaborative effort, we made the results publicly available, while hoping for more citizen science in the world. However, our own project remained unknown to the civilization until we had the luck to recruit a talented and dedicated community manager: Andreea.
Due to her coordination and the relentless effort of the entire team, we started establishing an online presence and managed to attract a couple of thousand players to our games. They played the Quantum Moves games roughly 300,000 times. It is the systematic analysis of these and the comparison to the results of state-of-the-art computer algorithms that has formed the bulk of work for the scientific team of the project over the past couple of years. The results are pretty exciting and we really look forward to giving you all details in a few weeks…