How to use data repositories for sharing teaching materials

Published articles should be reproducible – that’s the ‘gold standard’ in academic research. But how about teaching? Should instructors share their syllabi, handouts, assignments, Rscripts and data with other teachers – and how?

Developing a university course is time-consuming. Most are required to upload all materials to uni-internal platforms for students. So why should a teacher put in extra effort to make ‘replication materials’ for a course available for other teachers, and how?

Why publish ‘replication materials’ for a course

• Visibility: Other teachers out there who google syllabi and topics will learn that you ‘cover’ a specific field. You are more visible in the community.
• CV extension: When you provide links to your teaching materials, you create an extension of your CV that demonstrates your teaching experience, software skills, and professionalism.
• Other instructors can be inspired by your concepts, slides, examples. Good practice would be that they mention on their slides or in their Rscripts where they got their ideas from. For example, with permission, I used graphs and an example by Chris Adolph, and give the source on my slides.
• Find errors: If other teachers are using e.g. your Rscript they might find mistakes and let you know.

So where should you upload your teaching materials? A personal webpage, or a data repository?

 

Personal webpage or repository?

There are pros and cons when you provide your teaching materials on faculty/personal webpage versus a data repository. On your webpage, you control all uploads easily, can decide on the layout, and use the teaching section as a demonstration of your portfolio.

A repository, on the other hand, will be a lasting record of your teaching activity. People won’t have to search for your website. When you change domains or uni, the materials will still be available. Many repositories tell you how often your syllabus, Rcode or tutorials were downloaded. What I also find intersting, is that a data repository can provide an ‘official’ citation for your materials. Gary King describes how teachers can make use of a the Dataverse Network (open to all researchers) and increase visibility at the same time.

“A dataverse created for teaching is analogous to a syllabus, which is also one of many possible alternative views of an existing set of scholarly material. And just as with the increasingly common practice of establishing online archives of syllabi, a dataverse created for teaching can also be shared with others, if the instructor chooses to do so. This facility thus enables the scholarly and teaching community to capture some of the efforts of individuals so that we might all improve our teaching, and it also rewards individual teachers by allowing them to publish their dataverses within a Dataverse Network and to receive additional visibility and recognition for doing so. We have even developed a standard way of giving a formal citation to a dataverse so recognition for this type of teaching or curator contribution can be maximized as well.” (King, An Introduction to the Dataverse Network as an Infrastructure for Data Sharing)

This is how a citation can look like at the dataverse:

 

I use a mix of both

I started with a teaching section on my personal webpage. For each course, this includes data, Rscripts, slides and assignments (e.g. for Regression in R). Unfortunately, I have no idea how many people actually use these materials – and that includes my own students (do they really download the slides for exam preparation?).

When starting the Cambridge Replication Workshop I encouraged students to upload their final assignment, a replication of a paper, to our dataverse. This gave me the idea that I should upload the course materials there as well. This way I can see if they are downloaded and how often. Since this is a relatively new course, I was quite curious to see if there is interest. Plus, it would be great if other instructurs started replication courses as well, and they might want to see the materials. So my personal webpage now contains a link to the dataverse for all class materials instead of providing a download from my own page.

Finally, I myself modelled the course after Gary King’s course which includes replication, and based on a syllabus by Victoria Stodden at Columbia University, who kindly gave me advice on this. This was an additional incentive to share my adapted teaching materials as well.

 

Open teaching

While this post is about sharing syllabi, Rscripts, handouts and other class materials between teachers, there’s a general culture of open teaching for students, for example you can get videos and podcasts at Cambridge, Berkeley, MIT, the Khan Academy, Coursera. It goes without saying that sharing your syllabi, Rscripts, handouts etc. benefits students as well. It is sometimes a quicker way to find literature, Rcode or information (rather than watching a video). I regularly scan Rscripts and materials by other teachers and give my students links to their webpages (e.g. stats methods at Princeton). I should also note that I exclude copyrighted material and literature from my teaching material package when I share it openly.

 

Your opinion?

I would be interested what others think about distribution of teaching materials, and if anyone uses public repositories for that purpose?

 

Links

Syllabi are currently collected in various archives and on instructors’ webpages:

Society for Political Methodology Syllabi Repository

Miller Center of Public Affairs Syllabi Collection

Quantitative Methods Teaching Materials Archive at Oxford

Growing collection of Rcode for Introductory courses in R around Cambridge UK