ETH ORD Program

Authors
Affiliations

Lars Schöbitz

Global Health Engineering, ETH Zurich

Prof. Dr. Elizabeth Tilley

Global Health Engineering, ETH Zurich

Dr. Matthias Bannert

KOF Konjunkturforschungsstelle, ETH Zurich

This is a copy of the proposal as it was submitted to, and supported by the Open Research Data Program of the ETH Board. The original version is stored in this GitHub repository.

Proposal full title

Open WASH data by building Open Science Competencies and Community

Proposal Acronym: openwashdata

Background and Motivation

The Water Sanitation and Hygiene (WASH) Sector

The goal of the Water, Sanitation, and Hygiene (WASH) sector is to reduce the burden of disease by providing contaminant water in sufficient quantities, safe toilets that are hygienic and affordable, as well as solid waste management to reduce the spread of vectors (e.g. rats), especially within urban areas. Practitioners in the field produce several types of data: 1. Geodata (e.g. the location of a water pump along with data on its functionality); 2. Household data (e.g. demographic data related to income and a family’s willingness to purchase a toilet); 3. Operational data (e.g. the number of trucks and the type of waste they deliver to a dumpsite); or 4. Experimental data (e.g. the concentration of pathogens that die as a function of chlorine added to a water sample). All of these data are important to design programs and policies, and of course publish papers. However, despite being technical experts, professionals working in the WASH sector are not well-versed in the computational tools for research data management, including data analysis, visualization, or communication. They have not been exposed to sharing data following FAIR data principles and would not know where to start [1]. Typically, research projects in the WASH sector all have a common data management and publication workflow:

  1. Primary data is collected using survey tools or spreadsheet-based data collection tools
  2. Secondary data is shared by email, or using cloud-based tools file sharing tools
  3. Data is stored in spreadsheet-based software
  4. Data analysis is performed in spreadsheet-based software or in scripted proprietary software
  5. Results are copied and pasted into documents written in office word processing software
  6. Reports are exported to PDF documents
  7. PDF reports are published on funders websites, or in an academic publication which may be hidden behind a paywall making it difficult for practitioners without specialized subscriptions to access it
  8. Data remains unpublished or may be linked to an academic publication as an appendix
  9. If data is published, then in non-machine readable formats without metadata or documentation

Due to this lack of computational competency and/or data literacy, ORD practices that allow (research) results to be disseminated, including reproducible publications, data, software, hardware and protocols, are not employed. There are initiatives that have started to establish data standards, but they are either limited to one specific topic (e.g. Water Point Data Exchange [2]) or are lacking ORD practices themselves as recommendations are not turned into reusable formats (e.g. [3]). Further, none of these initiatives address the gaps in skills and competencies that are needed to employ standards. When ORD frameworks exist, then they are top-down initiatives by larger international donors or development banks [4], rather than grassroots approaches and community-driven initiatives.

The benefits for academics to apply ORD practices to their work are well researched and understood [5], but are not as clear for professionals that are working at the policy, implementation or evaluation components within a WASH project. As a result, data remains unpublished (or published not following FAIR principles [1]), collaboration is stifled, resources are wasted, and the wheel is constantly reinvented at extensive cost to donors and project beneficiaries. Further, the lack of competencies can make it difficult to establish partnerships between researchers applying ORD practices and partners that are not [6].

Data Science @ Global Health Engineering

One of the first people hired by Global Health Engineering once the position was created, was an Open Science Specialist. As a natural extension, we designed and now offer an undergraduate course for engineers which is the first of its kind at ETH: Research Beyond the Lab – Open Science and Research Methods for a Global Engineer (141-8102-00L, FS2022) [7]. In the first half of the course we teach the methods required to conduct real-word data collection and in parallel, we teach students about open science principles with a focus on: (1) Collaboration and version control with Git and GitHub, and (2) Research data management (i.e. open licenses, documentation and metadata). Students work in groups of three to design their own data collection tools (i.e. questionnaire, data collection tools for waste measurement) and collect their own primary data in collaboration with Entsorgung und Recycling Zurich (ERZ). In the second half of the course, we teach data science skills using R and the RStudio IDE. Lectures follow teaching methods of participatory live coding, peer instruction, and pair programming, with the majority of time spent on practising the learned competencies on their own collected data. At the end of the course, every group submits one group report contained in one git repository that was created collaboratively using git and GitHub. This course was the first step in developing a cohort of students that could then take on larger research, apply reproducible workflows and ORD practices to tasks within our group or elsewhere.

For practitioners, a “Building Communities to engage in ORD practices” workshop has been hosted three times; 181 email contacts were gathered through Zoom registrations. The workshop itself and all its materials are published under open licenses (CC-BY) [8]. A Slack Channel was established which 30 participants joined, and a Twitter handle (https://twitter.com/washdata) was started to share news about the further development. Of those that participated in the three workshops, 56 filled out a questionnaire after the workshop which helped us to gather an understanding about their starting point and interest in the topic. The results show that:

  • more than half of the participants had at least some experience using a programming language, while also about half had no experience using R statistical software
  • Nearly three-quarters of the participants store data in MS Excel Worksheets
  • Non-Governmental Organisations and Academia made up 60% of the participants
  • Only one participant expressed that there would be no interest in a 12-week course on data science

We asked participants about their perceived barriers for participation, where fees and time availability were perceived as barriers for the majority of participants. Therefore, we are pursing this Explore Call to design free online teaching material to follow up on these novel, self-funded offerings.

Problem Statement

The Global Health Engineering group at ETH Zurich works at the nexus of WASH research, anti-colonial development, and as a tool to achieve the latter two, open science. Yet, having worked internationally in the WASH sector as researchers, teachers, and consultants, we have seen how many resources are wasted in producing data, how much data is lost before it is made accessible, and how unmanageable the data that is accessible, is. This is equally true for colleagues in the Global South as it is for colleagues in Switzerland. We also know how intimidating data science can seem either because of the technology, the culture (which appears to many as being mostly young, white, and male), or the fear of failure.

The goal of this program therefore, is to develop a) a community of international and local students, and practitioners who are empowered to engage in open science concepts in order to b) create a platform to host WASH-related data which can be accessed and analyzed openly while c) researching the technical, pedagogical, and political factors that will allow us to expand and replicate the program to become the most comprehensive WASH hosting and training entity in open science.

ORD Project plan

This Project consists of 5 main Work Packages: WP0: Infrastructure design and installation; WP1: Mobilize community; WP2: Mobilize data; WP3: Data cleaning; and WP4: Data sharing and data publishing, which are elaborated in detail below.

WP0: Infrastructure Design and Installation

Goal: The Goal of WP0 is to prototype a technical foundation for the ORD toolchain at ETH in a a way that allows teaching and scientific collaboration to continue beyond the explore project’s funding.

RQ 0.1: What are the costs and requirements for ORD tools to be hosted and maintained within the ETH domain?

Activity 0.1: Collaborate with ETH IT Services to establish Virtual Machines and RStudio Server (open source version) for 50 learners

Activity 0.2: Evaluate the advantages and disadvantages of three open source tools for communication (Slack, Discord, Matrix Protocol)

Activity 0.3: Compare hosting costs and technical overhead (maintenance) of RStudio Cloud by RStudio PBC with Virtual Machines hosted at ETH Zurich.

During our upstream working package 0, we address costs and requirements for hosting a full fledged ORD process within the ETH domain. In the process, we will collaborate with ETH IT services to establish an open source backed ecosystem for a milestone of 50 learners. Infrastructure automation, e.g., through ansible automation platform is crucial to our approach: Scripted setups allow to reproduce our setup, disseminate our learnings and to foster ORD practices within ETH and beyond. Our work in WP0 is not limited to computing and version control infrastructure, we also plan to invest into communication tools that suit our approach of interactive collaboration in-person and online.

Based on our experience and previous work that lead to our submission, we intend to use open source software that is commonly applied throughout modern data science and statistics classes [9], including: R and RStudio Server for programming; Quarto scientific and technical publishing system for authoring [10]; Git and GitHub for version control and collaboration; Matrix protocol for real-time communication ; Mailing list server for regular updates directly to email.

Together with ETH IT services we will not only evaluate the best options to self-host our current explore project, but also make sure our virtual infrastructure can be extended or migrated to other providers easily. The above ecosystem will be used for teaching, collaboration and community building beyond this project and contribute to the greater understanding of how open source software can be integrated into ETH infrastructure, teaching and learning. Given the heterogeneity of the WASH community, a non-propietary is crucial for global adaption of a technical framework.

WP1: Mobilize Community

Goal: Create and grow an international network of data-curious practitioners and students who may be interested in pursuing one or more of the subsequent activities.

RQ 1.1: What are the demographic and competence characteristics associated with engagement as a function of communication channel/format?

Activity 1.1: Advertising the opportunities to contribute to the openwashdata community

Engagement 1.1: Read email, post, message and click to find information about the openwashdata program

RQ 1.2: What are the expressed and observed successes and challenges associated with live-coding events and how do these translate into program retention outcomes?

Activity 1.2: Weekly live-coding event to engage early adopters and demonstrate the low-barrier of entry to the data-sceptical

Engagement 1.2.1: Watch a live-coding event

Engagement 1.2.2: Follow along live-coding event

In WP1, we will mobilize WASH practitioners to join and contribute to the openwashdata community. Specifically, we will engage with our vast network of WASH students and professionals around the world by promoting the opportunity to a) learn more about open science tools and practices for improving the accessibility of their own work and, for the less engaged, b) offer the opportunity to publish their data in an open format. Although option a) is the more sustainable and preferable option, we recognize that the culture, tools, and terminology around data can be intimidating. We therefore hope to create as low a barrier to entry as possible, and hopefully, through small, consistent steps, move even the most reluctant, paper-loving researcher, through introductory steps, building their confidence and interest in the process.

We will engage through social media, email lists, professional networks (combined 30 years’ experience in WASH) and the established communication tools (combined 30 years’ experience in WASH). We will advertise the opportunities available to become a data sharer, data cleaner, or data publisher and communicator , and how to gain recognition (through permanent Digital Object Identifiers), and improve ORD practices, which, we anticipate will be attractive to the broader network, in exactly that order.

A weekly live coding stream using Twitch and Twitter will be one of our first entry points. Using the hashtags #TidyTuesday #openwashdata, we will advertise the weekly events in which an untidy (in our context unstructured data set) is shared and community members are challenged to tidy the dataset and produce visualisations and other data communication products. The live stream will use participatory live coding as a teaching technique, and participants can use the infrastructure provided (WP0: Prototyping ORD tools) to follow along the live coding [11,12]. These events are not designed to convert die-hard pen-and-paper practitioners into GitHub stars; rather, they are an opportunity to create low-stress environments for learning (no testing, no need to show results), identify community leaders who want to take on a more important role in dissemination, and share the magic of a well-structured data set.

There is an increasing understanding of the relevance of communities for open science and teaching [13] and we can see an increased pace at which they are being started for academics [14–16]. As active members of these communities, we will leverage the materials and experiences that already exist to build the openwashdata community. We will contribute back by sharing our experiences with the communities through open channels (blogs, communication tools, open access journals).

WP2: Data sharing

Goal: Solicit and receive interest in providing data to be published openly; develop an easy-to-use data-collection protocol and tool; receive data from interested partners.

RQ 2.1: What are the factors that predict willingness to share data and/or engage in self-directed data sharing

Activity 2.1: Create a GitHub repository template for submission of data, and documentation and metadata that follows general ORD standards.

Engagement 2.1: Watch video/ read documentation outlining data submission procedure

Activity 2.2: Reach out to participants identified in WP1 as well as pre-identified colleagues who have, but have not made their research data open

Engagement 2.2.1: Submit data unassisted.

Engagement 2.2.3: Engage with program team to assist with submission

RQ 2.2: What are the observed and stated challenges associated with transferring data to a GitHub repository?

Activity 2.2: Produce a suite of learning materials (videos tutorials, printed instructions, support-group platorms, Wikis) that enable interested parties to easily share research data

In WP2, we will identify community members that are open to share their data with the community to become “Data sharers”. We established the openwashdata GitHub organisation with a GitHub repository template (https://github.com/openwashdata/openwashdata) as the entry point for data submission.

As a first activity, contributors will create an account on https://github.com and will share their GitHub username with the project team who will add them to the openwashdata GitHub organisation. The project team will then use the openwashdata template to create a new public repository in the openwashdata GitHub organisation with push access to the GitHub username of the contributor. The contributor uses the GitHub Graphical User Interface (GUI) to add their raw data to the ~/openwashdata/data/raw_data/ directory and commit the changes to capture their submission as a contribution to the repository. As another activity, they are asked to create a persistent digital identifier on https://orcid.org/ (an ORCID iD) and add their family names, given names, and ORCID iD to the CITATION.CFF in the root directory by editing the file using the GitHub GUI.

These are the minimum required steps to contribute data to the community. Instead of receiving data by email, we ensure that contributors are exposed to version control and collaboration with GitHub and introduce the topic of open and reproducible workflows as an importance ORD practice [17]. We also ensure that each contributor can be credited for their submission appropriately downstream of the data sharing.

The highly technical nature of the industry standard reproducibility tools presents challenges for novices. Therefore, we will use applied examples, guided instructions and lots of practice to teach reproducibility as part of the data sharing process [18].

Data will be shared come by community members engaged in WP1 as well as from pre-identified sources. First, the unpublished GHE data is not currently accessible. Second, the data that was generated during Prof. E. Tilley’s 5-year tenure at the University of Malawi, Blantyre. As a research group within the ETH domain Eawag/Sandec’s research produces scientific knowledge that is paramount for the WASH sector. However, until this point none of this research has followed ORD practices and little of the highly valuable data that is generated as part of projects is published. As both applicants have previously worked at Eawag/Sandec and have long lasting partnerships, we will engage with current and former Eawag/Sandec colleagues as a third source of guaranteed data for this WP.

WP3: Data cleaning

Goal: Facilitate user-directed data cleaning (including documentation and metadata) to produce a first set of analysis-ready open data sets

RQ 3.1: What is the interest of openwashdata community members to contribute to data cleaning activities?

Activity 3.1: Invite and encourage openwashdata community members to contribute to data cleaning activities.

Engagement 3.1: Read email, post, message and click to find information about the data cleaning activities

RQ 3.2: What are the most common and most difficult-to-remedy data cleaning issues among openwashdata community members?

Activity 3.2: Categorise provided data sources into rubrics of “effort for cleaning” and “impact of clean data” to prioritise which data to invest in first.

RQ 3.3: What is the program adherence rate among openwashdata community members between previous WPs and WP3; and between the different modes of engagement within WP3?

Activity 3.3: Teach a 7-week synchronous online course (3 hours per week) for openwashdata community members to learn the competencies needed to clean submitted data.

Engagement 3.3.1: Attend at least one week of training

Engagement 3.3.2: Complete all 7 weeks of the online course

Activity 3.4: Host two online hackathons where selected data sets are cleaned by participants.

Engagement 3.4: Attend a hackathon

Activity 3.5: Host a 2-day workshop in-person workshop at a sector relevant conference with an element of capacity building

Engagement 3.5: Attend 2-day workshop

The journey for a “Data sharer” which started in WP2 can end there or continue into the role of a “Data cleaner” that actively supports WP3 either by using their data cleaning competencies (e.g. hired Research Assisstant) or by obtaining the competencies after participating in the learning events offered as part of this WP3. At this stage, other community members can also become “Data cleaners” without having submitted their own data in WP2.

As submitted data will vary in its effort to clean and the perceived impact of clean data, we will categorise the submitted datasets and use a decision support mechanism to decide which data sets are prioritized for data cleaning. Shared data will largely be unstructured and we will follow advice by [19] who suggest a minimal structure to share data for analysis. The GitHub repositories created from the template as part of WP2 will be used at this stage to add documentation and metadata in the form of machine-readable codebooks describing each data file. The project consortium brings in the expertise of the swissdata project, which aims at improving the machine readability of publicly available data (https://github.com/swissdata/demo) which follows similar goals to this project. We will use open material to teach open standards commonly used for reproducibility [20] and established guides to build a better basis for formatting data in spreadsheets [21] and how to get them ready for sharing [22].

Data cleaning processes will make up a large part of the workload for this project. Without the support of the community, it will be unfeasible to clean all submitted data sets within its duration. Rather than ticking a box next to a number for cleaned datasets, our goal is to motivate the community to engage in the data cleaning processes themselves. We do that through teaching activities, such as a free 7-week online course, hackathons, and workshops at conferences. These teaching activities will provide community members with the necessary competencies to clean data themselves and we will have enough resources from WP2 to use as examples. Beyond this, we hope to create enough interest in the value of the collected data sets to also engage with the greater reproducibility communities and R user groups in Switzerland (e.g. Swiss Reproducibility Network) and internationally (ReproHack, R4DS Community, AfricaR R user group, R-Ladies, Minority R Users, The Turing Way, etc.).

WP4: Data publishing and communication

Goal: Increase reach and impact of data by making it citable and creating data communication products

RQ 4.1: How does engagement with, and citation of data, increase by providing it following FAIR principles for data sharing and providing user-friendly mechanisms for discovery?

Activity 4.1: Publish GitHub repository containing data and code in general-purpose long-term archiving repository Zenodo.

Activity 4.2: Prepare data communication products using published data to generate new insights and results.

Activity 4.3: Prepare R Data Packages for published data sets with the most promising reuse potential

Activity 4.4: Aggregate published data sets into websites that are data warehouses and data catalogues

Where WP2 ensures that data is shared and WP3 brings it into an analysis-ready format, WP4 adds the layer of data publishing and communication. We understand data publishing as archiving of data in a long-term repository with curated metadata verified by the community [19], and data communication as the process of using analysis-ready data generate new insights and communicating them through articles, presentations, interactive documents, websites, and books. Once again, community members that have become “data sharers” and “data cleaners” can become “data publishers and communicators” to learn the final steps of the process of making their data accessible following FAIR data principles [1].

The GitHub repository that was created for the individual data set is archived and published on Zenodo, a general-purpose open repository operated by CERN and recommended by the the Swiss National Science Foundation (SNSF) as a data repository for long-term archiving and for making it citeable [23]. The publication process is straight-forward as Zenodo has a GitHub integration and the required documentation and metadata has already been generated as part of WP2.

To highlight the potential of publishing analysis-ready data openly, we will use the quarto framework to produce websites that highlight the research products. The use of quarto allows others to apply their knowledge within the R programming language that we teach, but also with Python, Julia, and Observable (JavaScript), making this the most inclusive and comprehensive data communication framework that exists.

The project aims to produce several websites, but at least, a (1) Global Health Engineering data warehouse, which highlights self-archived publications, data, notebooks, and further analysis, (2) openwashdata data catalogue, which provides a website with a searchable table of all data sets that are submitted, cleaned, and published by the openwashdata community. (3) R Data Packages and Research Compendia for the most promising data sets that were published [24].

Impact

Sustainability

Extremely. Although this is the first phase and highly experimental, it has the potential to generate not only a significant amount of data but also much needed knowledge about the barriers, needs, and challenges associated with moving a very decentralized, traditional discipline forward in it’s understanding and appreciation of Open Science principles and ORD practices. We are confident that we will identify leaders, both at ETH and internationally, who will be crucial to extending and scaling up future iterations of this work. By opening the door and seeing who comes in, we can identify the people, organizations, and programs that are interested in building on or replicating this work in the future either through their own networks, as student projects, or through formal funding mechanisms.

Addressing a critical problem

Despite the fact that hardware and software is more expensive, harder to access, and more difficult to support across much of the Global South, students and researchers have become increasingly dependent on commercial products that are oftentimes corrupted, pirated, or unstable. At the same time, the global community is demanding increasing quantities of data against which payments, punishments, and progress are judged. Researchers without the skills necessary to meet the increasingly open scicnece requirements of journals will be left with fewer publishing options. Empowering the WASH sector with the skills and data necessary to work openly is the only way to achieve the SDGs.

Collaborative approach

Our proposal is not only open to all research disciplines, thematic areas, geographical locations, and levels of expertise with respect to data literacy, computational competencies, or statistics knowledge, but we encourage and rely on diversity. To illustrate they diversity we expect, we have designed four learner personas (Table 1). Writing learner personas for teaching programming is a concept established by [25] and is a crucial step in understanding the needs and expectations of the participants.

Overview of learner personas designed for openwashdata community.
persona in brief domain knowledge programming knowledge contribution motivation
Palesa Pit emptying business owner that is tired of others asking her for her business data. competent novice low
Yua PhD student that wants to use her programming expertise to support the community. novice expert high
Mandla Master’s student who wants to learn how to use R for data analysis and git version control. competent competent high
Asim Senior Researcher with a few years left to retirement who wants to share his career’s worth of data with as little effort as possible. expert novice low

Palesa is leading a pit latrine emptying business with 20 employees and 6 vacuum trucks in Lusaka, Zambia. She has finished high school and did an apprenticeship in business administration. She is also a representative for the Sanitation Workers Association of Sub-Saharan Africa. She knows how to setup a data table using spreadsheet based software. She tracks all her business information using this software, but has never had any formal introduction into data organisation in spreadsheets. When asked to share to data, she shares the relevant worksheet by email. She is constantly approached by researchers, consultants, and public servants about sharing data on the amounts of types of faecal sludge removed from sanitation systems. While she is willing to share information, she is frustrated by and tired of follow up questions and the number of requests. She wants to learn how to give people access to her data at a central location, so that everyone can access the data from there. She is building a house for her family of five with the revenues of her business. She juggling many responsibilities and does not have any time to participate in formal training.

Yua is a Japanese PhD student in Computer Science at ETH Zurich. She is an active contributor to open source software packages and has a passiong for Open Science principles. She is proficient in half a dozen programming languages and uses R and Python for Data Science tasks. She has an R package for data visualisation on GitHub, which has received more than 500 stars. Her best friend works in the WASH sector and she has learned a fair bit about it. Her great experience and expertise in data cleaning enable her to decode data structures with minimal metadata. She wants use her technical skills to contribute to open communities and use her experience to support others in becoming more proficient in using computational tools. The openwashdata community provides the ideal starting point for her to do that. She also hopes to use the wealth of data that is opened by the community to advance her Computer Science PhD research on barriers to computational reproducibility. She is introverted and prefers to communicate with people online.

Mandla is in his final year of a Master of Science in Urban Planning at Makerere University in Kampala, Uganda. He is supporting Senior Researchers at the university in data collection for WASH related projects and is highly motivated to learn modern data science tools for his data analysis tasks. He did well in his high school math classes, and has watched some free online courses on using R for data analysis. He is struggling with file management and has recently lost three months’ worth of data, as his hard drive was corrupted. He knows that there is a better way to analyse and manage data, but he hasn’t found the right material and motivation that teaches him the tools that he needs. He wants to participate in 12-week course that teaches him the foundational knowledge and wants to apply the learned skills to support the openwashdata community in their mission to publish open data. He doesn’t have money to buy a laptop. He uses the infrastructure at the university, but during the recent student strike at the university the computer lab was damaged and is currently closed.

Asim has been working at a research institute in Aachen, Germany for 25 years. He holds a PhD and is a group leader with a focus on solid waste management research. Throughout his career, he has supported more than 100 students and research assistants in their early career research stages. He has never written a line of code, but is highly proficient in using MS Excel. He has a particular fear to share data with other researchers or alongside journal publications, because he believes that others could judge his practices or may even use his data to write their own publications. He is very particular about workflows, file and data management. During the offboarding process of each employee, he has ensured that generated research data stored on a long-term archive server, following his guidelines for directory structure and documentation. He wants to share this data with the openwashcommunity under the condition that he and every researcher that was involved in creating the data gets appropriate credit for their work. He has three more years until retirement is reluctant to learn any new tools and feels very uncomfortable with anything outside his normal working environment. He needs someone to show him step by step how to share he research data with the community.

Work Packages and milestones

The following Table 2 shows a basic gantt chart against the four work packages, including program activities and community engagement of the four defined learner personas. Column “Lead” abbreviations: MB = Dr. Matthias Bannert. LS = Lars Schöbitz. SA = Scientific Assisstant. The following Table 3 is a list of research questions associated with each of the Work Packages and related activities in Table 2. Any publications derived from this program will be published as open access material, following ORD practices and Open Science standards for computational reproducibility and sharing of data and code under FAIR principles.

Table 2: https://docs.google.com/spreadsheets/d/1T3NV75vVCeSRcN6FYghOrVFyWWi4WjQT5e4tqwp8jAQ/edit#gid=0

Table 3: https://docs.google.com/spreadsheets/d/105ZJL6bE-T8-fdm08e9sytFL9Vcj05NBbKnBcC2qkYk/edit#gid=0

Resources (including project costs)

Bibliography

1.
Wilkinson MD, Dumontier M, Aalbersberg IjJ, Appleton G, Axton M, Baak A, Blomberg N, Boiten J-W, da Silva Santos LB, Bourne PE, Bouwman J, Brookes AJ, Clark T, Crosas M, Dillo I, Dumon O, Edmunds S, Evelo CT, Finkers R, Gonzalez-Beltran A, Gray AJG, Groth P, Goble C, Grethe JS, Heringa J, ’t Hoen PAC, Hooft R, Kuhn T, Kok R, Kok J, Lusher SJ, Martone ME, Mons A, Packer AL, Persson B, Rocca-Serra P, Roos M, van Schaik R, Sansone S-A, Schultes E, Sengstag T, Slater T, Strawn G, Swertz MA, Thompson M, van der Lei J, van Mulligen E, Velterop J, Waagmeester A, Wittenburg P, Wolstencroft K, Zhao J, Mons B. The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data. 2016 Mar;3(1):160018.
2.
WPdx. WPdx The Water Point Data Exchange is the global platform for sharing water point data. 2022.
3.
Gather. Recommendations for the design of a data standard for sanitation data. 2020.
4.
Development Data Partnership - About. https://datapartnership.org/about/; 2022.
5.
McKiernan EC, Bourne PE, Brown CT, Buck S, Kenall A, Lin J, McDougall D, Nosek BA, Ram K, Soderberg CK, Spies JR, Thaney K, Updegrove A, Woo KH, Yarkoni T. How open science helps researchers succeed. Rodgers P, editor. eLife. 2016 Jul;5:e16800.
6.
GSMA. Innovative Data for Urban Planning: The Opportunities and Challenges of Public-Private Data Partnerships. Mobile for Development. 2021.
7.
Schöbitz L, Tilley E. Research Beyond the Lab Open Science and Research Methods for a Global Engineer. https://rbtl-fs22.github.io/website/; 2022.
8.
Schöbitz L. Data Science for WASH workshop hosted at the Colorado WASH Symposium 2021. 2021.
9.
Çetinkaya-Rundel M, Rundel C. Infrastructure and Tools for Teaching Computing Throughout the Statistical Curriculum. The American Statistician. 2018 Jan;72(1):58–65.
10.
Allaire JJ, Teague C, Scheidegger C, Xie Y, Dervieux C. Quarto. 2022.
11.
Brown NCC, Wilson G. Ten quick tips for teaching programming. PLOS Computational Biology. 2018 Apr;14(4):e1006023.
12.
Nederbragt A, Harris RM, Hill AP, Wilson G. Ten quick tips for teaching with participatory live coding. PLOS Computational Biology. 2020 Sep;16(9):e1008090.
13.
Azevedo F, Liu M, Pennington CR, Pownall M, Evans TR, Parsons S, Elsherif M, Micheli L, Westwood S, FORRT. Towards a culture of open scholarship: The role of pedagogical communities. MetaArXiv; 2021.
14.
Auer S, Haeltermann NA, Weissberger TL, Erlich JC, Susilaradeya D, Julkowska M, Gazda MA, Schwessinger B, Jadavji NM, Reproducibility for Everyone Team. A community-led initiative for training in reproducible research. Pérez Valle H, Rodgers P, editors. eLife. 2021 Jun;10:e64719.
15.
Button KS, Chambers CD, Lawrence N, Munafò MR. Grassroots Training for Reproducible Science: A Consortium-Based Approach to the Empirical Dissertation. Psychology Learning & Teaching. 2020 Mar;19(1):77–90.
16.
Kathawalla U-K, Silverstein P, Syed M. Easing Into Open Science: A Guide for Graduate Students and Their Advisors. PsyArXiv; 2020.
17.
Beckman MD, Çetinkaya-Rundel M, Horton NJ, Rundel CW, Sullivan AJ, Tackett M. Implementing Version Control With Git and GitHub as a Learning Objective in Statistics and Data Science Courses. Journal of Statistics and Data Science Education. 2021 Jan;29(sup1):S132–44.
18.
Ostblom J, Timbers T. Opinionated practices for teaching reproducibility: Motivation, guided instruction and practice. 2021 Sep;
19.
Tierney NJ, Ram K. A Realistic Guide to Making Data Available Alongside Code to Improve Reproducibility. arXiv:200211626 [cs] [Internet]. 2020 Feb; Available from: https://arxiv.org/abs/2002.11626
20.
Community TTW, Arnold B, Bowler L, Gibson S, Herterich P, Higman R, Krystalli A, Morley A, O’Reilly M, Whitaker K. The Turing Way: A Handbook for Reproducible Data Science. Zenodo; 2019.
21.
Broman KW, Woo KH. Data Organization in Spreadsheets. The American Statistician. 2018 Jan;72(1):2–10.
22.
Ellis SE, Leek JT. How to Share Data for Collaboration. The American Statistician. 2018 Jan;72(1):53–7.
23.
(SNSF) SNSF. Open Research Data: Which data repositories can be used? Swiss National Science Foundation (SNSF). https://www.snf.ch/en/WtezJ6qxuTRnSYgF/topic/undefined/en/WtezJ6qxuTRnSYgF/topic/;
24.
Marwick B, Boettiger C, Mullen L. Packaging Data Analytical Work Reproducibly Using R (and Friends). The American Statistician. 2018 Jan;72(1):80–8.
25.
Wilson G. Teaching Tech Together: How to Make Lessons That Work and Build a Teaching Community Around Them. Chapman & Hall/CRC Press; 2019.

Citation

BibTeX citation:
@online{schöbitz,
  author = {Schöbitz, Lars and Dr. Elizabeth Tilley, Prof. and Matthias
    Bannert, Dr.},
  title = {ETH {ORD} {Program}},
  url = {https://openwashdata.org/pages/gallery/proposal/},
  langid = {en}
}
For attribution, please cite this work as:
Schöbitz L, Dr. Elizabeth Tilley Prof, Matthias Bannert Dr. ETH ORD Program [Internet]. Available from: https://openwashdata.org/pages/gallery/proposal/