An introduction to data notes

This new section in Cogent Education has its origins nearly 10 years ago, when the move towards open data saw funders mandating that research data should be made more easily accessible by uploading them to repositories such as Figshare. Our discussions were greatly influenced by Colin Latchem (who sadly died in 2018) and resulted in a collection of data papers published in 2015 (see Rushby, 2015)

Data sharing makes the data which underpins empirical research papers available to the research community so that others can undertake their own analyses and build upon it for further research. Comparative research brings additional value and takes research to the next level globally. Data Notes bridge the gap between the research paper that gives the background to the research (the methodology, the analysis and, crucially, the novel findings and ground-breaking insightsand the data that has been stored in the repository. It allows the reader to understand the when, how and why data were collected, and what the data-product is.

To achieve this, the data note needs to include details of the dataset itself:

• Its location (the name of the repository) and its DOI

• Who created or collected the data and when

• The format of the data

• Any restrictions to its use

As with a normal research paper, it also needs to have a short abstract, an introduction, a short description of the research method and its limitations. The work of others should be acknowledged, and it should include relevant references. Reference should be made to the research papers which tell the full story and present the novel findings. Inevitably, there will some overlap in the contents of the research paper and the data note: the latter should include sufficient information so that it makes sense, but not try to replicate the research paper.

Open data confer many benefits, both to the research community and the individual researcher.

Much educational research is based on relatively small samples and consequently results in conclusions that can be challenged because they show little or no significant difference. Data sharing enables us to aggregate relatively small datasets and so make available large data corpuses. These may give us confidence that the effects are statistically significant.

Other benefits are that others can carry out their own analyses and perhaps think of other causes that could possibly explain—or contradict—the results. A fresh pair of eyes (or many fresh pairs) can sometimes provide a different perspective on a research problem. And lastly, but not least, data sharing can help to reduce unnecessary duplication of research.

On a less altruistic level, sharing data and publishing a supporting data note gives the researcher(s) three key benefits:

• An additional citable reference from their work leading to increased citations

• Greater discoverability for their work

• Opportunities for future collaboration with others who wish to extend the work.

Clearly, there are ethical issues that must be considered when sharing data. The data must be anonymous and the data note should be carefully written so that anonymity is preserved. However, there are hidden perils in large data sets that have apparently been anonymised, as Harvard University discovered in the aftermath of its of 1,700 Facebook profiles to explore the social-science data for “how friendships and interests evolve over time” (Zimmer, 2010). The creation of “big data sets” by aggregating data from different studies on the same subjects can make it possible for data miners to identify small clusters of subjects or even individuals.

In contrast to research papers, there is no well-established format for an educational data note and so, for the first year or two, we will all be finding our way to publish what is possible and useful. We look forward to working with the research community to provide a valuable resource.