Data Carpentry and Software Carpentry are committed to developing and supporting lessons in new domains, and to helping Carpentry community members do so as well. The development and maintenance of lessons requires significant resources, both up front and over the long term; this document outlines the requirements and procedures we have put in place to ensure success and sustainability.
1. How We Determine Which Lessons to Develop
We use the term lesson to mean a single teaching module, typically half a day in length; we use the term workshop to refer to an entire two-day short course, which typically comprises three or four lessons. The Carpentry community could create and maintain lessons and workshops on an almost infinite variety of topics. However, in the absence of infinite resources we must prioritize, and have chosen to do so according to the following criteria:
We are obligated to develop some workshops by the terms of supporting grants and partnership arrangements. Work on these necessarily takes precedence over work on other materials.
Our community repeatedly requests lessons on particular subjects, which may or may not be the same subjects advocated by our instructor community. Because our instructors are usually more computationally skilled, and computationally adventurous, than our intended audience, we place a higher priority on needs demonstrated by communities of learners.
The subject matter is stable and in widespread use (where “widespread” means “widely used within a particular domain or domains”, not just by the author and a few enthusiasts). As exciting as new tools may be, they are usually less stable than ones which are more established, and their longevity is usually less certain.
There is guaranteed support for initial development, delivery, and maintenance. (See the next section for more specifics on what we require.) This is not only a pragmatic requirement, but also confirms interest in the most concrete way.
No equivalent lessons exist—at least, none that do not assume prior knowledge, are living documents updated by their community, and use a hands-on teaching style.
The lesson materials will be freely available under a license that requires no more than attribution, such as the Creative Commons - Attribution license (CC-BY). We (strongly) prefer that lessons use open source tools, but do not require that. Similarly, we do not require that lessons be created from scratch, and in fact prefer contributors to recycle materials they have tested in the classroom so long as they meet licensing stipulations. We do require that lessons use our templates and teaching style.
Data Carpentry and Software Carpentry’s Steering Committees may make exceptions to these requirements in special cases, but such exceptions will be rare.
2. What New Lesson Proposals Need
As per point 4 above, proposals for new lessons or workshops must be backed by concrete statements of support for initial development, delivery, and maintenance. Such support should include:
One or two lead developer(s) who have made a credible commitment to develop and deliver the first complete version of the lesson. (Based on our experience with Data Carpentry’s lessons on genomics and geospatial data, a new two-day workshop requires 0.5 person-years of effort.) The actual effort may be less when existing material can be recycled, or when the lesson lead has significant prior experience with lesson development.
At least two trial sites, where early adopters with both domain expertise and experience teaching our general workshops have committed to delivering the first complete version of the lesson and providing detailed feedback.
Administrative support so that Data and Software Carpentry can coordinate work, assess outcomes, and report on results. Based on experience with developing Data Carpentry, we budget this at US$10,000 for a workshop US$5,000 for a half-day lesson. This cost may be incorporated into a partnership agreement with the sponsoring organization.
If a full two-day workshop is being developed, the budget must also include travel money for a hackathon at which early adopters (both those at trial sites and others) can work together to refine the first version based on early experience teaching it. This is usually a 3-day event for 10–15 people whose costs to participate are covered, at which the lead developer(s) will usually teach the entire material for early adopters to observe. Based on Data Carpentry’s experience, such a hackathon costs roughly US$20,000.
Our process for new lessons is:
The proposer contacts the Executive Director (ED) of either or both Data Carpentry and Software Carpentry by email to determine whether someone is already working on a lesson in this area, and if not, whether the criteria in Section 1 are satisfied. The EDs may introduce a proposer to others who have expressed an interest in the same materials so that they can coordinate efforts.
If one or both EDs express interest, the proposer works with the EDs and other members of the Software and/or Data Carpentry community to develop and submit a lesson development proposal. (See below for a template.)
Data Carpentry and/or Software Carpentry review the proposal. This may include soliciting opinions from community members and domain experts, or both. The review process is open: all comments will be visible to all community members. Community members who wish to join the proposal or support it in other ways may do so at this point.
If Data and/or Software Carpentry accept the proposal, they publish it on their website(s).
The proposers then begin work in an openly-accessible version control repository (e.g., GitHub), reporting progress monthly. They may adjust goals throughout the development process based on feedback from community members; substantial changes in goals must be approved by a representative of the appropriate Carpentry.
Once the lesson leads have a minimum viable lesson, they deliver it, and then provide feedback (both theirs and their learners’) to Data and/or Software Carpentry. This feedback must include items agreed to in the assessment plan incorporated into the Lesson Development Proposal.
The lesson is added to AMY (the web-based tool we use to manage the instructor database) so that instructors can indicate their ability and willingness to teach it.
After the lesson leads have revised the lesson based on early feedback, the Carpentries organize a hackathon to refine the material and introduce it to early adopters. This hackathon will usually include an on-site delivery of the material.
Early adopters (and the lesson leads) deliver the lesson at multiple sites, coalescing feedback and refining the lesson.
The lesson leads and Carpentry community members then publish the lesson on the Carpentry websites.
The lesson leads provide two hour-long online training sessions in which they discuss the motivation for the lesson, their experience developing it, and what they have learned from early deliveries.
The Carpentries solicit one or two volunteers to maintain the lesson, and the lesson enters the regular maintenance cycle.
Lesson developers are strongly encouraged to add their own institutions’ logos to the lessons. During development, lessons must use the Data Carpentry or Software Carpentry lesson template, but not their logos.
If no-one updates or teaches a lesson in a 12-month period, Data and Software Carpentry may decide to retire it. The community may continue to maintain a retired lesson, but the lesson will no longer be included in published releases or discussed during instructor training.
5. Lesson Development Proposal Template
Lesson proposals must include the items listed below. We expect that proposals will be developed in conjunction with Software and/or Data Carpentry, and that at least one of the proposers will have gone through Software and Data Carpentry instructor training, and be familiar with our teaching and lesson design methods.
1. Short Description
This should be at most 140 characters, e.g., “Tabular Data for Phylogeneticists” or “Data Management for Digital Humanities”.
2. Development Lead(s)s
Names and contact information of 1–2 lead developers, along with a one-paragraph biography for each outlining both their expertise in the domain and their teaching and curriculum development experience.
3. Target Audience
Who this lesson will target, e.g., “Graduate-level researchers in ecology”.
4. Lesson Outline
A brief description of what will be taught. Each half day of material should have:
- 3–6 concrete learning objectives.
- An end-of-lesson assessment exercise to demonstrate the skills participants have learned.
- A summary of the tools and data set(s) that will be used.
- A point-form learning plan.
- A brief comparison with existing open-access lessons on the subject.
5. Evidence of Need
Summarize evidence that researchers need this lesson. This summary may include links to online discussions (mailing lists, twitter, etc) or publications (e.g., descriptions of practices that are not yet widely adopted), results of surveys, etc.
6. Development Plan
Explain who is going to write and teach what, and when. The development plan must include a timeline that makes specific people responsible for specific lesson modules, commitments from specific sites to teach trial versions of the lesson, a date and location for a hackathon (if appropriate), etc. We recognize that this plan may change as lesson development progress, but the more specific it is, the more credible the proposal will be.
Explain who will support lesson development and how. If you have secured funding, attach details. If you have not, but intend to seek it, describe any planned or submitted funding requests. If the work is not being funded, explain how development and delivery will be supported.
As inclusive international organizations, we fully endorse instructors who wish to teach in the languages of their choice. While we would like to aid this by supporting translations of lessons into languages other than English, our experience is that doing this is comparable in effort to creating a lesson in the first place. In particular, maintaining a translated lesson as material evolves requires significant long-term commitment. We are therefore only able to aid such efforts if they are backed by the kind of resources described above for new lessons.