“Sharing your code and data is one of the things that make science science”
Dr Ana Martinovici is Assistant Professor of Marketing at Rotterdam School of Management (RSM). Her research is about how attention, and especially visual attention, reflects what people like, and what they are going to choose. In order to get to know our behaviour, she uses eye-tracking data. Martinovici is a strong advocate of openness and transparency and tries to make her research as open as possible.
What is your research about?
“I’m using eye-tracking data to know exactly what people are looking at, on a screen. For example, if you shop at an online retailer, we look at how your eyes move over the screen, how you choose between different options, and I can predict what you are going to buy before you actually put the item in the basket. Whether it is text or photos, the way your eyes move, tells me something about what you are interested in.”
In order to manipulate customers choices?
“Ideally not. That is where the ethical use of data comes in. Obviously, websites try to influence your decision. They don’t use eye-tracking data, luckily, but they do have all your clicks and therefore know a lot about your behaviour. My goal is not to make webshops able to sell more things that people don’t need.”
What is the goal of the use of eye-tracking data?
“I’m interested in the fundamental processes that take place when people are choosing. The link between what happens in your mind and your actions.”
You use open science, can you explain in what way and to what extend?
“Look at my usual Zoom-background, it is the GitHub logo! GitHub is a platform where not only researchers, but also software developers, and basically everyone who writes code, can share their code. It is useful for researchers because it allows version control. If I share my code now, I also share all the previous versions. People can go back in time and see how a project has evolved. This is also useful for myself because if I need the analyses I did last year, I can very easily restore them. I don’t have to search through old files on my computer.
I use the GitHub logo as Zoom background in all online meetings, online talks and seminars. It is an easy way to promote GitHub and therefore open science.”
Why do you find it important more researchers use open science?
“Sharing your code and data is one of the things that make science science. I find it frustrating when I see papers that are methodological, so they offer a method, but they don’t share the code that allows you to implement this method. As researchers, we should be open and transparent towards each other and towards the general public. Especially in the Netherlands where we are basically paid by the public.”
“Being open in science doesn’t mean that you have to share everything without legal or ethical considerations”
But you also make mistakes, what is the point of sharing the errors?
“I’m aware of the fact that there’s no code without bugs or errors. Of course, I hope there are no errors in my work. But if there are, it is better that people find them sooner rather than later.”
Does it also require a willingness to look at your own mistakes?
“Yes, and you have to overcome the fear of someone finding out you’ve done something silly. You have to accept that errors happen. But at the same time: this openness forces you to write better code because you are aware of the fact that other people will read it.”
What was your main reason to dive into the development?
“I did my PhD in Tilburg right after the Stapel-case. Being open was what you had to do, as the university wanted to avoid such a situation from happening again. I started using GitHub as a PhD student. There was a surprise benefit: I had my office-PC and a laptop at home. By using the GitHub platform, I could work on the same project everywhere without having to use USB-sticks or Dropbox. I can basically work on any device, anywhere, and I will not lose data if my laptop breaks down. Another benefit: I share teaching materials and I find teaching materials on GitHub as well.”
You cannot share for example eye-tracking data with everybody, right?
“That is how I got also interested in privacy, data protection and the ethical use of data – I also teach about privacy. Of course, there are limitations in what data you can share. Being open in science doesn’t mean that you have to share everything without legal or ethical considerations.
For one of my projects, my co-authors and I are using eye-tracking data we cannot share because of a Non-Disclosure Agreement. In order to open up the research, I generated a synthetic dataset. We say: this is simulated data. But it still allows readers to check our code, and to get more or less the same results as we have in the paper. In this way, we are open as much as possible.”
What would you recommend your fellow researchers to start with?
“I started using open science tools in 2014 and I'm still learning, it is a long process. For example, I did not yet preregister my studies. I recommend starting with small steps. First, create a GitHub account. Start with one project.”
What is the biggest benefit of working with open science?
“You sleep better at night. I know all my data is saved somewhere. And it is open, I’m doing my best to show other people that I’m honest in my research and findings. That is at least one less thing to worry about, in this messy world.”