A snapshot of the academic research culture in 2023 and how it might be improved

Octopus (octopus.ac) is a new research publishing platform with the goal of reforming how research is conducted, shared, and reviewed. It is different from other open research platforms by breaking down the reporting of research into eight distinct elements: 

• Research problem
• Rationale/Hypothesis
• Methods
• Results
• Analysis
• Interpretation
• Real world application
• Peer review of any of the above

This platform encourages the publication of research in smaller increments, with a lower barrier to entry compared to the traditional journal article format. In addition to being free-of-charge for authors, Octopus is distinct from other platforms because it requires published items to be linked to one another. For example, a method proposed by one researcher could be linked to a hypothesis (or multiple hypotheses) published by another.

Ultimately, Octopus aims to reform research culture in five ways: 

1. Reducing barriers and hierarchies that restrict research sharing, leading to more meritocratic recognition (especially for early career researchers and specialists)
2. Reducing pressures that can lead to questionable research practices (QRPs)
3. Addressing factors that lead to certain topics and types of research being favoured
4. Reducing bias in the assessment of publications and their quality
5. Encouraging more openly collaborative ways of thinking and working – division of labour and open critique of all parts of the research process

To understand whether Octopus is achieving these aims, and if its offering could be improved, we conducted a baseline evaluation of the academic research and publishing ecosystem. To do so, we employed three methods in order to achieve a broad understanding of research culture across different fields. Implementation details for the three approaches are described in the appendix, and summarised here: 

• Literature review - To avoid duplicating past work, and incorporate previous findings into our conclusions, we conducted a literature review on published studies on research culture relevant to the five aims of Octopus. This included informal searches, and keyword-based systematic searches on major publication databases.
• Semi-structured interviews - We invited researchers across the social, natural, and applied sciences to 1-hour interviews. These comprised open-ended questions regarding how they perceive the culture in which they work, specifically with regards to division of labour, sharing practices, giving and receiving critique and credit, and views on their careers.
• Online survey - We conducted an online survey to gather input from a greater diversity and number of researchers. It contained questions regarding overall research culture, barriers to sharing, division of labour, and factors affecting research assessment.

Our evaluation revealed a wide variety of barriers to more open sharing of research. While some are related to perceived or experienced biases based on personal characteristics such as gender or inequitable access to support, most result from a research culture that primarily assesses achievement and quality through traditional, peer-reviewed papers. This focus, and the resulting competition, encourages researchers to hide their work at least until a traditional journal paper is published. In some situations, these pressures lead to questionable research practices, such as data manipulation to achieve an "interesting" or statistically significant result more likely to appeal to a journal with higher impact metrics or perceived "impact". In general, open research practices are viewed as not beneficial, or even detrimental, to job security and career advancement. This is especially true given competing demands and the need for academics to prioritise their time on outputs that count in assessments that they are subject to.