This data collection consists of 27 participant information sheets and 23 consent forms freely available on-line. Materials were collected following a comprehensive search for publicly available ethical materials from randomised control trials (RCTs) targeting cancer (2007-17), primarily by systematically searching key on-line databases and monograph series. These corpora, which are different, to our knowledge, than any existing collection of medical English, could further research on information provision for patients in RCTs specifically and in healthcare settings more generally, in addition to advancing the study of the language of written ethical documents. Secondary analyses of these data could be undertaken using techniques from corpus linguistics, computational linguistics, and/or discourse analysis, for example, to investigate the nature and complexity of the language used and/or broach participants’ understanding of ethical principles or preference for how different language functions are expressed. Obtaining informed consent is an ethical imperative when conducting research involving human participants. However, participants’ actual level of understanding is often difficult and impractical to assess in operational research. One setting where the stakes for understanding are high due to the potential consequences of research participation is randomised controlled trials (RCTs), which test the effectiveness and safety of medical treatments. However, ethics committees' gatekeeping mechanisms often mean that legalese is mandated in consent forms, which can work against patients’ understanding. The goal of this text-based study was, therefore, to build and analyse a corpus of patient information sheets (PIS) and consent forms (CF) from RCTs conducted in the UK.

We drew on systematic review methodology to select sources to include in our corpora using the following criteria: Randomised controlled trials (RCTs) in progress or completed after 2007 (timeframe for inclusion: January 2007 to July 2017) targeting cancer in adult patients. All ethical materials needed to be publicly available online at the time of the data search for inclusion in the corpora. Nested studies not directly testing cancer interventions and emergency interventions were excluded. We used the search terms ‘Randomised Controlled Trials’ AND ‘Cancer’ as a Health Research Classification System Category, a UK Clinical Research Collaboration Category, or keyword using the following databases: (1) National Institute for Health Research (NIHR) Journals Library, (2) Europe PubMed Central for RCTs funded by Cancer Research UK, Prostate Cancer UK, Academy of Medical Sciences, Breast Cancer Now, Breast Cancer Campaign, or Dunhill Medical Trust, (3) Research for Patient Benefit (RfPB) database of funded studies, (4) Medical Research Council (MRC) database of funded studies; (5) Medline; (6) Embase, and (7) Cochrane Central Register of Controlled Trials. Following abstract screening, we proceeded to full-text screening and checked data saturation against Google Scholar entries and 'cancer' RCTs on the Online Resource for Recruitment research in Clinical triAls (ORRCA) database. The original files were converted from .doc or .pdf format to .txt format and anonymised, with separate files of collated ethical materials deposited for each genre (i.e., one .txt master file for all PIS and another for all CF, with citation information given for each RCT from which the ethical material was drawn in these master files).