[ Tools for Researchers] November 7, 2019
Why social scientists should engage early in the research life cycle
Research in the social sciences can be a linear process of data collection, analysis, publication that ends with dissemination. However, in practice it can also be a non-linear cyclical process, especially as new forms of digital communication allow ideas and findings to be shared and receive feedback at different stages throughout a research project. In this post Michelle Kuepper, Katie Metzler and Daniela Duca highlight the benefits that social scientists might find in engaging early with the research life cycle and the growing array of digital tools that are enabling researchers to do so.
The push for opening up the research process is picking up steam. Plan S, publisher mandates for sharing data alongside journal articles, and the rising popularity of preprint servers indicate a fundamental shift in the way that research is disseminated and a growing interest in understanding and representing research in a broader – end to end – fashion.
An appreciation for pre-published research has become well-established in other fields, notably physics, maths, computer science and the biomedical sciences. However, with the exception of economics, the early sharing of research has not taken hold in the social sciences, or humanities. Notably, arXiv, a preprint repository for physics set up in 1991, which currently hosts 1.5 million documents, hit the 10 thousand submission rate per annum after just three years. SocArXiv, set up in 2016 for social science research, has 4,500 documents.
This is a somewhat crude comparison, as there are about three times as many physicists as there are social scientists, meaning these numbers could in fact be more promising than they initially appear. However, they hint at a degree of disciplinary difference, for instance the need in the natural sciences to establish the priority of research findings. This is not to say the social sciences could not also benefit from adopting more open research practices. Specifically, social research stands to benefit in three ways; making the development of research more efficient, increasing the recognition of research, and ensuring the quality of research.
If you have ever published an academic book or paper, then you will know the research lifecycle is long and complex, often taking months or even years for research to go from ideation to publication. The vast majority of findings generated along the way are restricted to the offline world – effectively hidden in conference posters, computer files and field notes. This means that for the most influential and generative period of a research project, discussion of findings and ideas is confined to a relatively small institutional peer group, with wider discussions only taking place post-publication.
This doesn’t have to be the case. By openly sharing findings throughout the research process, insights can be drawn from a far wider group of peers, which can be used to hone and improve research projects. This reflects the way in which early research is traditionally shared at conferences, but by making these types of outputs available digitally it increases the chances that these early research outputs will reach a more diverse and relevant audience.
Finding and cultivating these engaged audiences is the first step on the road to finding potential collaborators that can join your project and contribute different knowledge and skillsets. This extends to feedback and collaboration from outside the academy, which if incorporated early can make research more impactful and even provide access to valuable sources of social or commercial data. Ultimately, a research article may be the most refined statement of a research project, but this doesn’t mean that the various points along the road to publication weren’t useful.
Finding new collaborators, developing a wide network, and increasing the reach of research is an important part of research, but it is especially important for early-career researchers looking to build a reputation when they don’t yet have a set of published articles under their belt. In the UK, the Research Excellence Framework has already taken one step towards recognizing the importance of early-stage research by accepting preprints as research outputs. There is also a growing awareness that research assessment should extend to an ever growing range of outputs beyond journal articles.
In a 2018 study in the Netherlands, researchers found that winning a postdoctoral research grant will inevitably lead to more funding in the future and better success in earning tenure, even for researchers with similar backgrounds, and similar levels of publications. Research into researcher productivity further shows that success is randomly distributed, making it something of a numbers game. The more meaningful ways you can provide routes into your research, such as by sharing different outputs like conference posters, the more likely you are to make the connections that render your work impactful.
Take for example, this hot-off the press preprint on predicting the success of startups, which although it is has yet to be peer reviewed and published, has already been cited. Attempts to measure the overall impact of pre-published research and other research outputs are in their infancy. However, for individual researchers the ability to publish and in particular assign DOIs (Digital Object Identifiers) to these outputs, provides a range of insights into the early impact of research.
Finally, pre-registering research and sharing data throughout the research process can help improve the reproducibility of quantitative studies, a particularly challenging topic in the social sciences. This is one reason many publishers have introduced mandates requiring authors to submit their supporting data alongside a final paper, and in many disciplines, especially psychology, preregistering research is becoming the norm. Not having access to detailed methods and data is one reason commonly cited for poor reproducibility rates, and sharing supporting data encourages researchers to engage in more rigorous research practices.
While qualitative data in the form of interviews and field-notes could be beneficial if shared, the conclusions derived from them are inseparable from the subjective experience of data-gathering and interpretations of the individual researcher. Due to their sensitive nature, these materials are often confidential. In fields like economics and political science, much of the data, if reliable in the first place, is historical and cannot be reproduced easily under controlled conditions. Nonetheless, having a clearly defined procedure and interpretive framework overseen by other researchers could improve the validity of these studies and avoid ‘HARKing’ (Hypothesising After Results are Known).
Sharing and posting a variety of files with a comprehensive and useable set of metadata (for each) can be a burden for researchers, especially when grant applications and final paper reviews are priorities. However, platforms for sharing early-stage research are increasingly interoperable and better integrated within the scholarly communication cycle than you might think. We have already mentioned DOIs, but platforms like the Open Science Framework also allow you to collaborate with peers while you share data and decide which parts of the project to make public. Likewise, tools like Morressier aren’t an extra burden for researchers, as they can also be used to submit posts and abstracts to the conference organizers.
As the importance of early-stage research continues to gain traction, social scientists can already take the first steps towards embracing this development and begin sharing their pre-published research. This will enable both individual researchers and research communities to start reaping the various benefits that an open approach to the research lifecycle can bring.
About the authors
Michelle Kuepper is the Head of Communications at Morressier, the platform for early-stage research that aims to accelerate scientific breakthroughs by making previously hidden findings discoverable. She has a background in science and technology journalism and tweets at @mischabk
Katie Metzler is Associate VP of Product Innovation and strategic lead for the SAGE Research Methods product portfolio, which includes Cases, Datasets and Video. She leads the SAGE Ocean initiative which aims to equip every social scientist with the skills and tools they need to do the social science research of the future. As part of this initiative, she led the development of SAGE Campus, a suite of online courses teaching research methods and data science skills to social scientists. She is also responsible for the product management function supporting SAGE’s digital products for the library market, including SAGE Video, Data Planet, SAGE Stats, SAGE Knowledge, CQ Researcher and Business Cases. Follow Katie on Twitter @KMetzlerSAGE
Daniela Duca works on new products within SAGE Ocean, collaborating with startups to help them bring their tools to market. She is also a visual artist, with experience in financial technology and has a PhD in innovation management. You can connect with Daniela on Twitter.
This article originally appeared on the LSE Impact Blog.