Making open science work for you: Practical tips for incorporating open science practices into your everyday workflow

• In Event: 2-151 - Step by Step: Improving best practices for research with children

Integrative Statement

In recent years, there has been a push toward making research more transparent and reproducible. This effort includes, for example, preregistering studies, analyzing data and writing papers in R, and sharing data, code, and materials. Recent workshops and symposia have focused on training researchers in these tools and practices, but such training is only effective if researchers incorporate this knowledge into their everyday workflow. Many researchers, however, perceive barriers to using open science practices: “Where do I start?!” “I don’t have time!” “If I don’t do it all at once, it’s not worth it!” “How do these practices work for developmental psychology?” The aim of this talk is to begin addressing these limiting beliefs, provide researchers with concrete suggestions for how to start incorporating open science practices into their workflow, and to highlight the benefits of doing so.
One common concern is that open science practices will take too much time. However, in many cases these practices largely rearrange where in the research process activities occur rather than increasing overall time spent on the project. For example, pre-registering a study requires planning details such as the sample size, exclusion criteria, and data analysis prior to data collection - activities which would occur anyway, but traditionally later in the research process. Moreover, pre-registration gives additional benefits including enhancing research transparency, facilitating collaboration with co-authors, providing a clear record of research decisions, and increasing the overall quality of the planned research. As another example, learning a new programming language such as R will require time at the outset. However, many researchers report that this upfront investment ultimately saves time and increases accuracy. Because R is a scripted language, code for analysis or figures can be easily re-run when additional data are gathered, or adapted and re-used for similar projects. Tools like R Markdown eliminate the tedious and error-prone task of copying the value of test statistics from the output of statistical software to a results section.
Another common belief is that “doing Open Science” is all or nothing. Researchers might feel overwhelmed at the thought of immediately incorporating all of these practices into their workflow. However, most researchers who have adopted open science practices have done so incrementally. Easier places to start include pre-registering a straight-forward study (for example a replication), or sharing data on the Open Science Framework website. This experience can serve as a starting point, enabling the researcher to begin with the practices with which they feel most comfortable and/or believe are most important for improving the transparency and replicability of their research.
In this talk I intend to provide examples from my own experience incorporating open science practices into my everyday workflow and suggestions elicited from other developmental researchers at various career stages who regularly integrate such practices into their work. In this way, I hope to address the above-described concerns by providing suggestions for manageable, incremental changes that researchers can make in their everyday research process to begin aligning them more with the principles of Open Science.

Author

• Jessica Elizabeth Kosie, University of Oregon
  Presenting Author