Introduction

Replication vs Reproducible Research

A basic tenet in science is the ability to replicate the results of any experiment. Replication verifies results; however, research papers often lack the detail required for independent replication. Many attempts at replicating the results of well-known scientific studies have failed in a variety of disciplines.

Replicating studies with new data is expensive, for this reason computationally reproducible research is recommended as a way to assess the validity and rigor of scientific results. Research is considered reproducible when others can reproduce results of a study with only the original data, code, conditions and documentation.

Reproducible research has most of the advantages of replicating studies without the financial and time burden associated with collecting new data. It is the by-product of care and attention to detail throughout the research process and seen as a minimum standard that all researchers should strive for.

Video: Why reproducible research?

Benefits of reproducible research

  • Easier to recall and explain work to collaborators, supervisors and reviewers
  • Faster to modify analyses and figures
  • Facilitates quick reconfiguration of research tasks for new projects with similar tasks
  • It can increase the quality and speed of peer review
  • Can increase citations and expanded research impact through the ability to access and cite the project code and data
  • Enhances opportunities for research methods training
  • Reproducible research is a strong indicator of rigour, trustworthiness and transparency

Some scenarios to consider:

  1. A key person from your research team has disappeard (family, personal emergency or no longer contactable).
    • Could you continue your work?
    • Do you know where your data is stored?
    • Could you keep working effectively for 1 month? 1 year?
  2. You lose your laptop bag containing your all your external hard drives.
    • Could you continue your work?
    • Is your data backed up? Encrypted?
  3. Someone has published contradicting results to your published paper
    • Can your provide your data and methods?
  4. A research partner organization believes your sensitive data has been breached or made available to a third party.
    • Can you show the steps taken and measures put in place to avoid this?

Work through the following lessons for practical solutions and good practice ideas.

NEXT -->