I’ve been trying to convince Chris to set up a blog for the lab. A lab blog can take many forms: it can be public or private, formal or informal, edited by a single person or several people and so on. There’s quite a few potential benefits to getting a lab blog, for instance:

• Publicity. Public blogs can generate a significant amount of traffic, which may be helpful in recruiting subjects, grad students and postdocs. For an investigator that is starting out, the publicity could be helpful in getting his name out there. It can also be very good for disseminating new papers.
• Writing practice. In the case of a collaborative blog, in which grad students and postdocs write posts, it gives them a nice opportunity to practice their writing skills. Chris commented that “students who come in have NO experience writing”. He was talking especially about people with backgrounds in math or physics or engineering that wrote a handful of papers for undergrad classes. It’s hard to get better at something by doing it a few times a year.
• Outreach and community. As scientists we arguably have a duty to communicate to the public and to other researchers. A blog is an easy way to talk about science to the general public and build professional relationships with other researchers. This is especially good if you shun conferences for whatever reason (Personally, networking at SFN or even trying to learn stuff there is an exercise in futility, and I’m sure I’m not alone).

I emailed Jonathan Pillow and Liam Paninski who have their own lab blogs and asked about their experiences. Dr. Pillow has a public blog that contains different types of posts. The majority of the content consists of summaries of the weekly journal club, written by different grad students and postdocs. Jonathan says that he “likes the idea of having a public record of the papers / ideas we discuss in our lab meeting, and also simply as a way for ourselves to keep track of “where we’ve been” intellectually over the past year.”. I presume the summaries also give students writing practice. The rest of the posts contain announcements concerning participation in upcoming conferences and new papers. Jonathan says that it’s been a good experiment so far.

Dr. Paninski’s blog focuses on announcements of upcoming group meetings, which include discussions of their work as well as reviews of other papers. It’s more of a private affair than the Pillow lab blog. Liam says that “now we just use it mainly to keep track of things we’re reading and thinking about.”. He also mentions that very little time has been spent on the blog and that it’s been a positive experience so far.

It doesn’t have to be a timesink

The main problem with creating a blog is to keep it updated, and that could be a timesink if you’re doing it wrong. What you need is a source of continuous content and assign responsibilities to other people. If you’re a PI, you don’t actually need to write content yourself; just have the minions do it (hey, it’s writing practice!). An easy way of getting continuous content is to force people to do a short summary of what they discussed after your journal club and/or lab meetings. It could be the presenter, or it could be the person who presented the week before, say.

A large proportion of the content can be stuff that is non-peer-reviewed, but nevertheless useful to you and your labmates, and needs to be archived. Notice that Drs. Pillow and Paninski both mentioned the usefulness of a blog as a way of keeping track of things. I know that personally, I often search for some old stuff I posted, whether it’s to extract code or figure out how I solved a problem a long time ago. In that spirit, the following could make good content:

• Code
• Bugs in software that you figured out how to get around
• Important life lessons learned while buying/installing lab hardware
• Posters and presentations that you give at conferences
• Posters and presentations that you’ve seen at conferences
• Some new method/analysis that you tried and advantages/pitfalls
• Your ideas about a series of papers that you just read
• Grand-scheme-of-things talk about life, the universe, and everything
• Whatever that summer student did that was pretty cool but’ll never get published
• That appendix you had to cut out of the paper because it didn’t fit in

Most of this content, you already have, the blog just gives you an excuse to polish it enough so that you can understand what you did a year later. In addition, a lab blog can present ephemera, for example announce an upcoming presentation, presence at a conference, visits to other places, etc. Finally, you can announce and explain your recent papers.

Now, if you’d like to discuss your current research, and are paranoid about getting scooped, you can make the blog private. This is still consistent with the idea of using the blog as a means of archival. It won’t, however, give you the benefit of extra traffic to your website, but perhaps that doesn’t matter to you.

How to start a blog and get it “out there”

Although it’s possible to have a custom install of blogging software on a university server, it’s probably easier to simply sign up for a hosted blog. It takes 5 minutes. Two suggested platforms are WordPress (used by Dr. Pillow) and Blogger (used by Dr. Paninski). I’ve been told that Blogger stats are better, but I like WordPress in that it’s quite flexible content-wise; it parses and code, and you can attach to posts pdfs, word documents and PowerPoint presentations (not zip files, however). In both cases, it’s straightforward to make a collaborative blog; this entails having people create a user account at, say, WordPress.com, and then having the admin add the users as contributors to the lab blog.

If you’re unfamiliar with blogging, basically all platforms offer a WYSIWYG editor in which to enter posts, as well as facilities for uploading and making minor edits to images. While it is possible to change the look of the blog by using a different theme than the default (see xcorr with another theme for an idea of what’s possible), usually the blog look cannot be changed arbitrarily on hosted services (you have to pay extra), but honestly, who cares. It’s possible, in addition to posts, to have a few static pages, for example for an About page. Users can subscribe to your blog in a variety of ways, usually through the RSS feed; they can also subscribe through WordPress itself or email; you can also push updates to Twitter and Facebook. They can like posts and retweet and do all sorts of social media goodness. This is how you get return traffic. Finally, posts can be commented on. WordPress comes with a pretty good spam filter for comments.

If you’re interested in getting a decent amount of traffic, in addition to posting quality content, you may try commenting on other blogs, which is an easy way of making others aware of your existence. You can register at Nature Blogs, which will, if nothing else, increase your visibility to search engines. Finally, and most importantly, once you have a decent number of posts (say 10) you can become a part of researchblogging.org. The site aggregates blog posts directly discussing peer-reviewed research; it’s mostly read by other scientists and nerdy audiences. Basically, you add a bit of HTML after a post discussing a peer-reviewed paper or papers that serves as a bibliography readable by the researchblogging.org crawler. It’s a little bit of extra work, but you get a lot of traffic from it; I usually get about 100 hits the day after I post something to researchblogging. There’s no rule that states that you can’t discuss papers from your own lab, as long as you write a real blog post about it (that is, something less formal than a paper or preview, which perhaps includes background information or discusses open issues in a way that wouldn’t be possible in a scientific publication).