Thursday, June 26, 2014

Draft --> Writing Software with Analytics for Writers

Another item originally written for folks at Macmillan Education.

Traci Gardner, who contributes to Bits, a Bedford/St. Martin's blog for teachers at, tweeted a link to a ProfHacker entry by Konrad Lawson about Draft, a tool for collaborative writing, at ""  Lawson observes,
Draft is designed for drafting and collaborative writing of text. It is not a blogging platform or a live editing environment like Google Docs, and the documents created within Draft are not designed to have their final home there. As it functions now, you write, collaborate, edit, and then export or directly publish the documents to your cloud hosting service such as Dropbox, Evernote, or Google Drive, to a social platform such as Blogger, WordPress, Tumblr, and Twitter or, if you use the Chrome or Firefox additions, directly back into a text box in another browser tab. Get a quick overview of its features here: .

I want to draw attention here to the analytics ( features. Especially this idea, described by Draft's creator, Nathan Kontny:
One mistake I keep seeing people make, when they publish their writing, is that they don't pay enough attention to attributes that might affect how much traction that writing will get.  They'll publish 2000 word posts, when their audience would prefer 500. Or they publish on Friday night, when no one might be paying attention and Monday morning might be a better idea. I wanted to make this type of analysis a lot easier to understand, and help people, including myself, learn what makes our writing get more attention than other writing.

This is the kind of thing -- finding ways to give writers data they can use to make decisions about their writing -- that really allows the writer to adapt; it gives them agency. That's a distinct from the kind of adaptive learning that pushes a learner to one desired end, a preset end, and where the software varies content and activities based on performance for the learner to do until they reach the end, where the software is adapting, not the learner.

Note the kinds of things Kontny's trying to help with -- audience awareness, the ability for writers to see differences in drafts. A lot of online writing tools (, for example) can give statistics on word count, or in the case of Grammarly, the number of sentence level errors made. And there's some use in that some times, but giving writers information on what in their writing finds an audience, when it's best to post, what the reading level of their writing is (Draft uses the Flesch reading level.), and other information lets a writer see things over time. This intranet does some of that, by the way; it reports (click the "Content" tab and then filter to see your own content) what of your contributions have been viewed and how many times, commented on, liked, replied to. That's standard stuff in social networks and should be standard stuff in the learning spaces we make for students. What would up the value on that would be aggregation for totals of all those items, and a view that shows trends over time.

The other feature of Draft is the ability for a writer to compare drafts side by side, whether one's own writing or writing being co-edited by fellow writers offers a powerful tool as well. Teachers struggle to get writers to address global revision, what is sometimes called higher order revision. Students tend to focus on tweaking sentences here or there. So a tool like Grammarly or Word's spell checker/grammar checker draw attention to surface level issues. But a tool like Draft shows changes. That's the kind of thing we need in the writing tools we make for students and teachers. There needs to be a drafting tool that supports version control, to encourage writers to save as or upload drafts. And a tool that not only lets writers compare drafts, but that highlights changes. And not only highlighting changes, but it should also tell writers what percentage of the draft is changed: how many new words were added, how many new sentences were created, new paragraphs, how many deletions were made. The idea would be a tool that can summarize larger revision from minor editing. And then something that aggregates that so over assignments and drafts, a writer sees trends.

Giving writers and writing teachers these views -- evidenced based insights into writing and revision processes -- gives them tools for articulating change and describing growth. A teacher can ask for a new draft and can tell the writer that she wants to see at least a 25% increase in new ideas. Peer reviewers can return to see which of their feedback ideas lead to bigger revisions by the writer, which would help reviewers see the value of their work and why it matters.

1 comment:

Nate Kontny said...

Thanks for reviewing those bits in Draft and spreading them! I appreciate that very much.