How can energy research practices be improved? A new approach emphasizing transparency, reproducibility and quality in a recently published Buildings & Cities peer-reviewed article explains what energy researchers can do to improve their research design, execution and writing.

A recent peer-reviewed Methods paper, published in Buildings & Cities, explores how practices could be improved in the conduct and publishing of energy research. Huebner, Fell & Watson (2021) highlight the importance of good quality research for applied energy studies, which often inform policy and can be of direct relevance in determining appropriate action for tackling and adapting to climate change. Given the urgency and gravity of this task, they argue that the rigour and quality of energy research should be subject to publishing practices similar to those used in the psychological and medical sciences. This is crucial reading for researchers beginning their careers, those already established, but also research councils and funders. In particular, these readers can gain insight into how to conduct and assess high quality research, and identify resources to support with the planning and implementation of a range of research approaches.

The authors focus on three dimensions of 'good research': transparency, reproducibility and quality and provide a clear process for thinking about how to implement this into research practices. In the paper, transparency is taken to mean making the essential components of research (for example, sampling and analytical approaches) visible to others. Reproducibility considers how information can be provided such that independent studies can seek to reproduce results. The authors' preferred definition of quality is less clear, but the implication is that this includes research which 'poses important questions, uses appropriate methods, assesses bias, and considers alternative explanations for findings' (p. 3; following a definition from the National Center for Dissemination of Disability).

Drawing on ideas and practices adapted from other scientific fields, particularly medicine and psychology, four specific tools are introduced for achieving transparency, reproducibility and quality:

• preregistration of analysis plans (pre-analysis plans)
• making data and code publicly available
• using preprints
• employing reporting guidelines  
  

Pre-analysis plans specify the study aims, types of data collection, and data analysis approach. The authors highlight the value of these for preventing the manipulation of research data to present favourable results. Reporting guidelines provide a means to support authors in determining what information to include in papers, for example sampling methods and recruitment processes. The authors suggest that preprints can be helpful for overcoming publishing bias (for example, where journals give preference to studies which demonstrate significant findings). Finally, the authors advocate for open data, or placing available data in an online repository. They highlight research which finds that research using open source data and code receives more citations than those publishing from closed data. A useful mantra, highlighted in the paper, is the European Commission's (2016) principle that research data should be 'as open as possible, as closed as necessary'.

The authors acknowledge that these different tools have varied degrees of relevance for distinct approaches. For example, exploratory research (where a strategy for analysis is not pre-determined) may be less amenable to a pre-analysis plan. However, with hyperlinks to pre-analysis plans and reporting guidelines for a variety of research approaches (including both qualitative and quantitative strategies), the article is itself a useful repository to work from when planning and publishing research. In addition, adopting these approaches will help to align energy research with accepted best practice in fields with rigorous checks on research quality, such as medicine.

The authors emphasise the need for 'more structural adjustments to the energy research ecosystem' (p.14). In particular, they highlight that some poor research and publishing practices are reinforced by the expectations of journals, for example prioritising the publication of research with significant results or confirming hypotheses, rather than research with 'negative' results.

Diverging from this norm, Buildings & Cities welcomes Methods articles (like that discussed here) which outline new techniques and critique existing approaches, and Replication articles which explicitly set out to test previous findings, validate existing data, and report results demonstrating 'failure'. With this Methods article, Huebner, Fell & Watson have opened up dialogue around research quality, and how tools used in the publishing process can create reassurances around this. The next task is to work out how to continue this debate such that the broad gamut of energy researchers can coalesce around a mutually agreed set of guidance.

Huebner, G. M.,Fell, M. J., & Watson, N. E. (2021). Improving energy research practices:guidance for transparency, reproducibility and quality. Buildings and Cities, 2(1), 1-20. DOI: http://doi.org/10.5334/bc.67