background-image: url("data:image/png;base64,#https://raw.githubusercontent.com/EstherPlomp/2022-PRES-Data/main/images/Esther.jpg") background-position: 80% 80% background-size: 50% ## Open Science: Data sharing Slides by [**Esther Plomp**](https://estherplomp.github.io/) @ TU Delft, Faculty of Applied Sciences License: CC-BY --- class: center, middle # Slides are available ### https://estherplomp.github.io/2022-PRES-Data/ --- class: inverse, center, middle # Research Data Management --- # Research data Any type of information that is collected, observed, or created, in the context of research -- - **Raw/primary data**: The data measured/recorded without manipulation -- - **Processed/secondary data**: Data that has been modified/processed for analysis -- **Data Management Plan (DMP)** to plan how to manage and share the data (see [The Turing Way for more information](https://the-turing-way.netlify.app/reproducible-research/rdm/rdm-dmp.html)) --- # Data Organisation  Image by [Allison Horst](https://github.com/allisonhorst/stats-illustrations) --- ## File naming - 20220113-PRES-Data-V001 - [8 step guide](https://resolver.caltech.edu/CaltechAUTHORS:20200601-161923247) on how to set up your file naming convention - [Presentation on file naming](https://speakerdeck.com/jennybc/how-to-name-files) - [Stanford’s best practices](https://library.stanford.edu/research/data-management-services/data-best-practices/best-practices-file-naming) -- ## Folder structure - Templates by [Colomb et al.](https://doi.org/10.5281/zenodo.4410128), [Nikola](http://nikola.me/folder_structure.html) and [Barbara Vreede](https://github.com/bvreede/good-enough-project) for [cookiecutter](https://github.com/cookiecutter/cookiecutter) - [Find Files Faster](https://zapier.com/blog/organize-files-folders/): How to Organize Files and Folders - [Data Management: File organisation](https://surfdrive.surf.nl/files/index.php/s/gJtISAQABapLvnI) by Christine Malinowski - [Videos on project structure](https://www.youtube.com/watch?v=u6MiDFvAs9w&list=PLRPB0ZzEYegPiBteC2dRn95TX9YefYFyy&index=2) by Danielle Navarro --- # Data Organisation .pull-left[ ] .pull-right[ ## Spreadsheets - [Spreadsheet organisation tips](https://surfdrive.surf.nl/files/index.php/s/9aNX69fsWyAnMI8) - [Broman and Woo 2018](https://doi.org/10.1080/00031305.2017.1375989) - [Wickham 2014](https://vita.had.co.nz/papers/tidy-data.html) - Use tools for data validation like [OpenRefine](https://openrefine.org/) ## Why? What could possibly go wrong? - [a lot](http://www.eusprig.org/horror-stories.htm) - See the [Twitterpost source](https://twitter.com/mdancho84/status/1426970168660529153)] ] --- # Data Documentation .pull-left[ ] .pull-right[ - (electronic) Labnotes - Readme files ([template](https://cornell.app.box.com/v/ReadmeTemplate)) - [Guide for data documentation](https://doi.org/10.5281/zenodo.1914401) - [Data Dictionary](https://uf-repro.github.io/data-organization/slides.html) - [Code Book](https://libguides.library.kent.edu/SPSS/Codebooks) #### More information - Book: Data Management for Researchers by Kristin Briney - [A Quick Guide to Organizing Computational Biology Projects](https://doi.org/10.1371/journal.pcbi.1000424) by William Noble - [Some Simple Guidelines for Effective Data Management](https://doi.org/10.1890/0012-9623-90.2.205) by Borer et al. ] .footnote[[Twitterpost source](https://twitter.com/thecrobe/status/1373590641012322306)] --- class: inverse, middle, center # Data Sharing --- # Open Data made freely available for use and re-use by anyone and everyone -- ✅ **Access** : Available (on the internet) to all on demand -- ♻️ **Reuse/distribution** : Provided under terms that permit reuse and redistribution -- ℹ️ **Transparency** : Providing information about data generation/collection -- 🌎 **Interoperability** : Interoperability with other data, machine readable formats -- 🤲 **Participation** : Everyone must be able to use, reuse and redistribute -- ➕ **Equity** : Data is not truly open if the research process is not open to all .footnote[ [#bropenscience is broken science](https://thepsychologist.bps.org.uk/volume-33/november-2020/bropenscience-broken-science) by Kirstie Whitaker and Olivia Guest [Open Science Beyond Open Access: For and with communities](https://doi.org/10.5281/zenodo.3946773)] --- # Not Open Data .pull-left[  ] .pull-right[ '[odds of obtaining the dataset] fell by 17% per year' ‘research data cannot be reliably preserved by individual researchers’ - [Vines et al. 2014](https://doi.org/10.1016/j.cub.2013.11.014) "We received no response to 41.3% of our data requests" - [Tedersoo et al. 2021](https://doi.org/10.1038/s41597-021-00981-0) ] .footnote[[Meme explanation](https://knowyourmeme.com/memes/agnes-harkness-winking) [Twitterpost source](https://twitter.com/MadS100tist/status/1366103674989277185)] --- # Open Data .pull-left[  ] .pull-right[ ## data repository online archive that curates research datasets and provides long-term access - Finalised datasets - ~10-15 years (Long term preservation) - Access - DOI = more citations/visibility - File format support [How can you make research data accessible?](https://www.software.ac.uk/how-can-you-make-research-data-accessible) by Esther Plomp] .footnote[Image by [Errant Science](https://twitter.com/ErrantScience/status/1251118457279647746)] --- # Why not supplemental materials? 🛂 **Data control**: cannot be updated -- 🌐 **Interoperability**: not available in all formats which makes it difficult to integrate and interact with the data -- ✅ **Availability**: Difficult to access if the article is behind the paywall (supplemental materials are not included in the DOI and therefore the links can also break!) -- 🏆 **Impact**: Data should be a primary research output -- 🏮 **Publisher requirements**: Some publishers recommend using a data repository instead -- [The Push to Replace Journal Supplements with Repositories](https://www.the-scientist.com/news-opinion/the-push-to-replace-journal-supplements-with-repositories--66296) --- # How to find a repository? - Check publications in your field - [FAIRsharing](https://fairsharing.org/) - [re3data](https://www.re3data.org/) General repositories: - [4TU.ResearchData](https://researchdata.4tu.nl/en/) - [Zenodo](https://zenodo.org/) --- # Data Licenses .pull-left[  ] .pull-right[ ## Data [Creative Commons License Chooser](https://chooser-beta.creativecommons.org/) ## Software [Choose an open source license](https://choosealicense.com/) ] .footnote[Image [Source](https://foter.com/blog/how-to-attribute-creative-commons-photos/): CC-BY-SA] --- # How to link your publication and data/code/protocols? - Publish the output before you publish the article OR - Reserve the DOI ## Use the DOI/citation in your publication Reference your data in the **Data Availability Statement** and the **References** .footnote[ [The Turing Way: Linking Research Objects](https://the-turing-way.netlify.app/communication/citable/citable-linking.html) ] --- # Publish or reserving a DOI ### [Zenodo](https://zenodo.org/) -> [Upload](https://zenodo.org/deposit/) -> [New Upload](https://zenodo.org/deposit/new)  --- # Linking with publication ### Data accessibility/within table (descriptions)  ### Data availability statements (at the end)  .footnote[ [Data accessibility source](https://doi.org/10.1016/j.dib.2021.107375); [Data availability statement source](https://doi.org/10.1002/ajpa.24059) ] --- # Linking with publication  Always check the dataset's readme file or metadata on how the contributors prefer to be cited! See [this document](https://surfdrive.surf.nl/files/index.php/f/5896617587) for more information about data/software citation. --- background-image: url("data:image/png;base64,#https://raw.githubusercontent.com/EstherPlomp/2022-PRES-Data/main/images/schmid-2021.jpg") background-position: 50% 50% background-size: 70% # Linking data/code/publication .footnote[ [Publication](https://doi.org/10.1038/s41565-021-00958-5) // [Data & Code](https://doi.org/10.5281/zenodo.5059802)] --- # Questions? -- # Discussion? -- ### “publication of data and codes should be mandatory" -- ### "All data (also raw data!) and code underlying the publications should be shared" --- class: center, middle # Thanks! Slides created via the R package [**xaringan**](https://github.com/yihui/xaringan).