Twitter activity at #PlantBio18 – Who, what, when, why?

Plant Biology 2018 (“A joint meeting of the American Society of Plant Biologists, Canadian Society of Plant Biologists/Societe Canadienne de Biologie Vegetale, and the International Society of Photosynthesis Research at the Palais des congrès de Montréal“) is over, and appears to have been record breaking in terms of number of attendees and sharing of the events on social media, particularly Twitter. For those of you who don’t use Twitter, it has become a popular medium for dissemination of the meeting “buzz”, both for those in attendance and those observing from afar. (More info on how and why people Tweet at conferences can be found here and here).

After the meeting ended, Nick Tomeo shared some data about the Twitter activity, and I’ve asked him to share how he did this analysis and what he learned about the Twitter habits of attendees.

By Nick Tomeo  @Tomeopaste

For the past few Plant Biology meetings I have compiled summary statistics and made plots to provide a synopsis of the meetings’ activity on Twitter. Especially for the past two meetings that I was not able to attend, this provided me a way to feel involved. These summaries are always exciting for many the folks that are active on Twitter. Some of us are even a bit competitive about our activity. I want to provide a short overview of how I access and analyze the tweet data for the meetings. If you are interested in a more thorough run through of how this all works, I have cleaned up the code I used to access and analyze the #PlantBio18 tweets. You can find the code along with data files containing all of the tweets on GitHub: https://github.com/Tomeopaste/PlantBio18.

Let’s take a look at the general overview of the tweet activity using the #PlantBio18 hashtag at this year’s Plant Biology meeting. The analysis and numbers here include a few days preceding and during the meeting (9 July to 23 July).

There were a total of 8,681 tweets from 1,241 different accounts. 2,386 tweets contained original information. 175 accounts tweeted at least five times, 110 tweeted at least 10 times, 51 tweeted at least 20 times, 18 tweeted more than 50 times, and 8 tweeted more than 100 times!

Twitter has an application programming interface (API) allowing you to query their tweet database and download all tweets matching your query specifications. The API makes collecting all tweets for a given hashtag relatively painless if you have a good means of interacting with it. Thankfully, there are interfaces in R: the twitteR and rtweet packages. This year I used rtweet to collect all of the #PlantBio18 tweets and cannot recommend it enough if you want to try this. To use the API, you need to get access granted directly by Twitter. Once you have access to the API, searching the tweets is as simple as providing an R function with the hashtag you are interested in. The data that is returned contains the tweets with that hashtag along with a treasure trove of metadata including the account, time, and geographic location of the tweet and how many retweets and favorites it received.

I used the timing of the tweets to summarize and plot the meeting’s activity by day and hour – see below for the circadian pattern of Tweets during the conference. Note that the vertical lines indicate midnight UTC, or 8 PM in Montréal where the conference was held. There are clear dips in Twitter activity at lunch and dinner time, when attendees are dining together rather than Tweeting!

Tweets by hour – lines indicate 8 PM conference time

I then aggregated all of the tweet metrics by account or Twitter handle. Using the aggregated data I made plots for individual user’s activity with the hashtag. The most active accounts were: @ASPB, @Biokid001 (ASPB President-Elect Rob Last), @itsbirdemic (Kevin Bird), @OshnGirl (Jennifer Robison), @plantae_org, @PlantTeaching (ASPB Features Editor Mary Williams), @RishiMasalia, @schwessinger (Benjamin Schwessinger). This year I plotted the users with the most tweet activity, the most original tweets, and the most impactful tweets.

·         Tweet activity includes all tweets, retweets, and quoted retweets. Topping the list as most active was Jennifer Robison, @OshnGirl.

Original tweets includes only tweets or quoted retweets. Benjamin Schwessinger @schwessinger was the most original.

 Impact was calculated as the sum of all favorites and retweets a user received divided by their total number of tweets. Mary Williams (@PlantTeaching) was the most impactful. (To help guard against extreme bias in the impact metric, only users with ≥15 tweets were included.)

Thanks to everyone at the meeting who participated on Twitter. I hope it enriched your experience of the meeting as much as enriched my following it.

Thanks for that Nick!

This year we also awarded prizes for the “Best Tweets” (judged by Plantae Fellows and friends). The runners-up were:

And

And the winner:

What does this all mean?

One great outcome of conference Twitter activity is that it helps to connect people who are Tweeting at the conference, which is especially valuable for first-time attendees (most conferences have an early “TweetUp” for just this reason).

Another outcome from all this Twitter activity is so that it makes it easy for others not attending to learn about the exciting science being presented (for example, who was invited to speak, and what the hot topics are), and provides good publicity to attract attendees to #PlantBio19 (San Jose California, August 3-7 2019).

Finally, Twitter is a great forum for continuing post-conference discussions. Speaking of which, Rob Last, who assumes the ASPB presidency in October, is eager to hear comments, suggestions, and insights about ways to ensure that ASPB supports all members of the broader community (members, authors, meeting attendees, and others), so please chime in using #ASPBforward or join this discussion.

Thanks finally to everyone who participated in the “conference within a conference” at Plant Bio 2018, and I look forward to breaking this year’s records at PlantBio19 with you!