Skip to main contentSkip to footer
A group of researchers sat around tables listening to a talk in a conference setting
Brighton & Sussex Medical School

Navigating open science while working with secondary data and large international teams

BSMS > Research > Navigating open science while working with secondary data and large international teams

Navigating open science while working with secondary data and large international teams

This case study focuses on making scientific research openly available, accessible, and reusable to 1) increase transparency and integrity, 2) promote collaboration and increase public value of the research. Specific practices discussed include pre-registration of analysis plans and open data/code. The case study describes applying these open practices to a large international consortium that has been running for over 20 years, comprising over 150 collaborators across 31 countries—the International Tobacco Control (ITC) Project. ITC aims to evaluate the impact of tobacco control policies and apply findings to inform national and international tobacco policies, such as packaging, warnings, and smoke-free legislation. ITC principally uses longitudinal and cross-sectional international surveys of adults who smoke and/or vape.

BACKGROUND IMAGE FOR PANEL

Experience 

I lead the England arm of the ITC Project. Since working with ITC since 2018, I have implemented open science practices when using ITC data. This includes pre-registering analysis plans on the Open Science Framework, as well as uploading code. ITC data are also available to researchers upon completion of a data request form and approval by country PIs, and the surveys and technical reports are published online by the ITC team. I have also led presentations to the ITC team on the benefits of incorporating open science practices into ITC.

 

Benefits

ITC produces high-quality, in-depth data across multiple countries. Sharing the surveys, technical reports, and analytic code can support colleagues in using and analysing the data. I have personally found that pre-registration of analysis plans is particularly beneficial when working with multiple co-authors (many ITC papers have 10+ authors) because, not only is it best practice, it also sets expectations at the start of the project and so can help mitigate challenges later. This is particularly helpful when supervising students or supporting early career researchers who are analysing ITC for the first time, as ITC is a comprehensive and complex survey so knowing specifically what outcomes and predictors will be included and how the data will be analysed up-front reduces challenges later down the line. Additional benefits of pre-registration include having written half of the paper (i.e., introduction, methods) before obtaining the data, journals look favourably upon it, and it gives you an additional output on your CV (particularly helpful for students and early-career researchers!).

 

Challenges

First, while open science practices are now commonplace in most UK universities, it is less of a priority in some other countries although this is changing. Some international colleagues are therefore less familiar with the importance of open science; however, the ITC team have been incredibly supportive of pre-registration and uploading code. Second, ITC is an established ongoing survey; as such, data analyses are secondary, and pre-registering survey design and sample size/power for surveys already fielded is not possible; however, the Open Science Framework has a pre-registration form specific to secondary data analysis that can be used. Third, while I advocate that students and early-career researchers at BSMS always pre-register analyses for ITC papers, such practices are challenging given time constraints and students’ lack of familiarity with the complexities of ITC data.

BACKGROUND IMAGE FOR PANEL

Lessons learned

Pre-register analyses when working with large teams, as it ensures everyone is on the same page before starting the project. Advocate for students to complete a pre-registration form to practice developing research questions and thinking through analyses, even if this is just for training purposes and not made publicly available. Don’t try and do too much open science at once — start small (e.g., with pre-registration) and then build up to broader practices (e.g., uploading code).