This document gives the latest content of my workshops about open science workflow. Please look into www.access2perspectives.com for a broader offer.

open science = (reproducibleanalysis + FAIRdata) x open

While funders, universities and publishers start to ask researchers to produce open FAIR data, and while open science is an objective of the European commission, the practice of open science and reproducible research is impeded by the lack of (good) training. My workshops are taking a constructivist approach to teach the digital and the non-digital skills needed to efficiently and effectively fulfill the novel standards in data management, data analysis and scientific reports. Participants are trusted to work on their own experiment, getting direct feedback on their practice and their flaws. In the end, participants will have created FAIR data, ready to be made open and analysed for one of their experiment.

Objectives

For participants: For labs:
Acquire data and project management skills required in academia and in the industry More collaboration in the lab: spot errors quickly
Save time and produce better data and reports Prepare your lab to the transition to open FIAR data and open science.

Most requested Workshop

1. Research data management: from planification to publication

2 days workshop focusing on data literacy and management (including data tidiness, raw data storage, metadata production, as well as open-material and methods), learning to plan ahead and save time on the long run.

In order to be able to share data in a useful format, one needs to plan it ahead. Data literacy and management are therefore becoming core competencies for scientists, and some universities are introducing data management course in their main portfolio. Indeed, the absence of data management plan before the start of the data acquisition leads to an enormous waste of time, and a little bit of digital data literacy can already make a huge difference.

Making sense of data is only possible with appropriate explanations. The file containing this information is called metadata, and designing the metadata to fulfill standards, allowing data reuse and interoperability is not as difficult as one think. Material and methods information should be included in the metadata, and knowledge about new tools and version controls becomes very handy there.

These advanced (digital and non-digital) skills are not only highly valued in academia, but also in the industry.

Constructivist approach

Participants will work in small group being either the data expert for their experiment, or a putative collaborator with little insights when working on others projects. They will experienced the difficulties in sharing their work and making sense of the work of others and face the problems of human and computer readability first hand. I will guide them in their learning, presenting key concepts of research data literacy and management.

The workshop is particularly suited for groups aiming at fostering collaboration between their members, because participants will explain their workflow and their data to their colleagues as part of the work.

Detailed Content

Data literacy

  • Raw/primary/derived data
  • Tidy spreadsheets, Excel pitfalls and the .csv format
  • From glossaries to ontologies: data interoperability
  • Metadata
  • Folder organisation
  • Workflow: archive before analyzing
  • Data publication (FAIR checklist), CC0 license
  • Basics of data analysis documentation

Version control

  • Version control principles
  • Where to apply version control
  • Git basics
  • Git intermediate

Material and methods

  • Material: RRID and key reagent open table
  • Methods and its publication

Data analysis and reproducibility

  • Derived data, version control, analysis documentation recap
  • Problems with excel
  • Advantages and pitfalls of using R (demo)
  • Rstudio suite: R, Rmarkdown, RShiny: reproducible reports (different levels)

Data analysis and statistics

  • sample size
  • Sampling error, p-hacking, multivariate analysis
  • Effect size and “significance”
  • Pre-registration

Experimental design

  • blindness, double-blindness
  • randomization and constant variables
  • control group design
  • neurogenetics

Trainer

Dr. Julien Colomb
experience:
 10 years as a researcher (neuro-biology)
 5 years in data management and open science

 
schillerpromenade 4
12049 Berlin

julien.colomb@fu-berlin.de

ORCID iD iconorcid.org/0000-0002-3127-5520