Wow! I’m a Moore Investigator in Data-Driven Discovery!

I am incredibly excited to announce that I am the recipient of one of the Moore Foundation’s Investigators in Data-Driven Discovery awards.

To quote Chris Mentzel, the Program Director of the Data-Driven Discovery Initiative:

Science is generating data at unprecedented volume, variety and velocity, but many areas of science don’t reward the kind of expertise needed to capitalize on this explosion of information. We are proud to recognize these outstanding scientists, and we hope these awards will help cultivate a new type of researcher and accelerate the use of interdisciplinary, data-driven science in academia.

I feel truly honored to have been selected. All the finalists that I met at the Moore Foundation in July were amazing as were all of the semi-finalists that I knew. I did not envy the folks making the final decisions.

So what will we be doing with this generous support from the Moore Foundation?

• Doing data-intensive prediction and forecasting in ecological systems: We’ll be focusing on population and community level forecasting as well as ecosystem level work where it interfaces with community level approaches. We’ll be using both process based ecological approaches with machine learning, with an emphasis on developing testable predictions and evaluating them with independent (out-of-sample) data. As part of this effort we’ll be making publicly available forecasts for large ecological datasets prior to the collection of the next round of data, following Brian McGill’s 6th P of Good Prediction (in fact we’ll be trying to follow all of his P’s as much as possible). There’s a lot of good work in this area and we’ll be building on it rather than reinventing any wheels.
• Increasing the emphasis on testable prediction and forecasting in ecology more broadly: Industry and other areas of science have improved their prediction/forecasting through competitions that provide data with held out values and challenge folks to see who can do the best job of predicting those values (most notable in Kaggle competitions). We’ll be helping put together something like this for ecology and hopefully integrating that with our advanced predictions to allow other folks to easily make predictions from their models public and have them evaluated automatically as new data is released.
• Tools for making data-intensive approaches to ecology easier: We’ll be continuing our efforts to make acquiring and working with ecological data easier. Our next big step is to make combining numerous ecological and environmental datasets easy so that researchers can focus on doing science rather than assembling data.
• Training: We’ll be helping build and grow Data Carpentry, a new training effort that is a sister project to Software Carpentry with a focus on data management, manipulation and analysis.

You can read the full proposal on figshare if you’re interested in more details. The slides from my talk at the finalists symposium are also available.

I’m very excited to be joined in this honor by my open science/computational training/data-intensive partner in crime C. Titus Brown(@ctitusbrown). I was also particularly thrilled to find out that I wasn’t the only investigator studying ecological systems. Laurel Larsen is in the Geography department at Berkeley and I can’t wait to interact with her more as we both leverage large amounts of ecological data to improve our understanding of ecological systems and our ability to forecast their states in the future. We are joined by astronomers, statisticians, computer scientists, and more. Check out the entire amazing group at the official Moore Foundation Investigators site and see the full press release for additional details about the program.

The award is being run through the University of Florida since we are in the process of relocating there, but I owe a huge dept of gratitude to the Biology Department and the Ecology Center at Utah State University for always supporting me while I spent time developing software, working on computational training initiatives, and generally building a data-intensive ecology program. Without their support I have no doubt that I wouldn’t be writing this blog post today.

Ernest Lab Ph.D opening at University of Florida

So here it is, the first of the positions we’ll be advertizing as part of our move to the University of Florida. The official ad is below, but a few comments first. The position is for a student to work with me, but for those who aren’t really familiar with our groups, it’s important to note that my group works closely with Ethan White’s lab (we provide desk space that mixes the labs together, we have a single group lab meeting,  etc). My group tends to attract people who like to do field work. Ethan’s tends to attract people who are more quantitatively or computationally inclined. We mix our groups because we believe that the divide that exists between quantitative and field-based approaches to ecology is bad for our science and that we need more people trained to serve as bridges between the quantitative and field-oriented worlds of ecology.

Here are some links to the papers my students have published from their dissertations to get a feel for what my students have intellectually gotten out of this environment:

Supp & Ernest (2014) Species-level and community-level responses to disturbance: a cross-community analysis

Macroecology Meeting in the Works

If you want more info, you should email one of the people who signed the below email (I’ve linked to their websites). I’m not an organizer, just a messenger!

Dear All

We are delighted to announce an upcoming joint meeting of the BES Macroecology Group, the GfO Macroecology Group, and the Center for Macroecology, Evolution & Climate (CMEC). The meeting will be hosted by CMEC in Copenhagen, Denmark during June 2015. We are sure it will provide an exciting opportunity for the members of these groups to share their latest research and ideas, and to initiate new collaborations in the relatively informal atmosphere consistent with the society group meetings. 

To help us find the best dates, length of meeting and a good estimate of participant numbers, we would appreciate it if you could spare a couple of minutes to fill out this very short survey: https://www.surveymonkey.com/s/KKFBMHY

Thanks very much 

The organising committee

CMEC: http://macroecology.ku.dk/

BES Macroecology: http://macroecologyuk.weebly.com/

GfO Macroecology: http://www.gfoe.org/en/gfoe-specialist-groups/macroecology.html

UPDATE: Fixed lots of broken links and a couple of typos

Organizing a Gender Balanced Conference

There is a lot of discussion on the internet about highly skewed speaker lists at symposia and conferences. For the past year, I’ve been co-organizing a small conference (~110 people) with Michael Angilletta where we’ve been practicing some of the approaches I developed and blogged about earlier for organizing a seminar series. However, in ecology we know that what works at small scales may not apply to larger scales. So, do I still think organizing a conference that is both strong on research and gender diversity is very doable? Read and find out.

But before I give you my thoughts, first, some stats and background. For this conference, we had a lot of moving pieces: Discussion Leaders who helped organize their sessions, Invited Speakers – both for long and short talks, a Mentoring Program for Young Scientists which involved selecting both mentors and mentees. In the end (I hope, I’m writing this about a week before the conference, so hopefully things don’t change drastically), we ended up with the following numbers for each of these parts.

Discussion Leaders: 5 men, 4 women

Invited Long-Talk Speakers (40 min talks): 9 men, 9 women

Invited Short Talk Speakers (12 min talks):9 men, 7 women

Students/Postdocs in Mentoring Program: 10 men, 10 women

Professors (all ranks) in Mentoring Program: 11 men, 9 women

So over all the slots that we invited people to fill, we have 53% men.

What did we do? Just like I talked about in my post about the seminar series, we generated a large pool of names. We started by making a big list of people to lead the various sessions at the conference and developed an invite list that was balanced. We then used our balanced group of Discussion Leaders to brainstorm potential speakers. Each Discussion Leader provided a list of people they thought would be excellent for their session. They were given detailed instructions about how to generate their list – diverse perspectives on their topic, diversity of taxa/ecosystems, including domestic and international scientists, and a reminder to be aware of the gender ratios of their list.

From those lists, Mike and I sat down and constructed our dream team speaker list – balancing research areas, topics and taxa/ecosystems, career stages, making sure we had some international representation, and keeping an eye on gender balance in the process. Then we set out to convince these people to come speak at the conference.

For the Mentoring Program, we ran an application process. We advertised on every social media outlet and listserv we could think of. Our pool of applicants was very gender balanced (23 women, 21 men). We selected 20 young scientists, equally split among male and female, again balancing across various dimensions of research & people diversity.

The Mentors and Short Talk Speakers were harder. Most of our Short Talk Speakers are students from the mentoring program but we had some slots leftover to fill. Both the mentor and short talk speakers needed to fill specific topic requirements either for the program or to overlap with already chosen mentees.

So what lessons did I learn?

Gender balancing a conference is hard, but not in the ways I generally heard about before I started. It is not harder to find female speakers – as long as you don’t restrict yourself to only senior female professors. There are lots of kickass women out there, but you need to embrace the fact that they are scattered across career stages. Women were not more likely to say no. I don’t know why we didn’t experience this commonly reported problem. Maybe it was because I was sending the invites. Are women more likely to say yes to another women asking? I also spent a lot of time sending personalized emails communicating why we thought they in particular would be a good fit for the conference and why we wanted them there (I did this for the men too). What was hard about it was spending the extra time sending personalized emails to communicate clearly why I was inviting them. Did those efforts make a difference? I really don’t know. You’ll have to ask the amazing scientists who said yes to our invites.

Developing at the get-go a diverse pool of people you would like involved is critical. This is another time intensive step. Crowd-sourcing this to our Discussion Leaders helped a lot. Many of them knew speakers (men and women) we hadn’t thought of. When we pooled all those suggestions, we had 123 suggestions for 16 speaking slots. That gave us a ton of flexibility when thinking about the program we wanted to create. It was also really handy when someone said no because all the brainstorming work had already been done. We could sit down with our list and come to consensus quickly on the next invite to send. We often saved up rejections to fill as a group, thus allowing us to manage the diversity better.

The more restricted the slot you’re trying to fill, the harder it is to get gender balance. If your need is 2 kickass people who work in general area X, then gender balance is easy. The more criteria you place or the fewer the number of slots (or both) the harder it gets. Need a senior researcher studying organism X on specific subtopic Y and need another senior researcher studying organisms Z on specific subtopic A in ecosystem Q? Yeah, both those slots are probably going to end up being men, just because of the numbers game. View your program creatively. Be willing to think about different ways people can fit into the program given the diversity of research you’re trying to cover and the multiple facets that everyone has in their research programs.

So my final thoughts on the matter? Making a gender balanced conference is not easy and because of the strong gender skew at the senior levels, it doesn’t just magically happen. It takes work, planning, creativity, and a great team of people helping you brainstorm names. But a 80:20 split in invited speakers is far from the grim ‘reality’ that some might think.

Pre-tenure Advice: Blocking out time for your research

As part of the Carnival that Prof-like Substance is organizing on Pre-tenure advice, I thought I’d throw in a piece of advice that anyone who asks me this question gets from me. Here it is:

Create a calendar and block out time for you.

Sounds simple, and honestly a little stupid, but it’s the best advice I can give. Why?

When you start your job, or a semester, your calendar is empty. You have oodles and oodles of time for you to work on your science. However, you will quickly find that there are a lot of demands on your time. Students want meetings. Faculty want meetings. Collaborators want meetings. Administrators want meetings. You name a type of person (and some you suspect might not be people at all) and they will want you to do something for them. Many of these things will seem quick. The most dangerous words are “This will only take 5 minutes” (It won’t. I promise you, it never does). Next thing you know you have a week chopped up into meetings with only 15 minutes here or there of ‘free time’. That’s now the free time you have to do all that science you were planning on. Trust me, in the 15 minutes you’ll have between meetings and things you have to do to prepare for meetings, you won’t feel like working on your science.

This is where the calendar comes in. Schedule blocks of time during each week for research. Once on your calendar, these blocks are sacrosanct. You wouldn’t cancel that meeting with the Department Chair to meet with Fred down the hall to talk about the seminar committee. Your research is in many ways more important for your future than the Department Chair, so don’t cancel on it.

Why do you need the calendar? Why can’t you just tell yourself that every Monday morning you will focus on research? Because it doesn’t work. When someone comes and asks for a meeting and you think about your week, Monday morning will be ‘free’ in your head. The easiest thing in the world is to fill that ’empty’ slot and now you’ve just broken up your research time into useless chunks. (Trust me, I’ve been there). On the other hand, if that time is already blocked out on your calendar as ‘busy’ then it’s a reminder that you have something you need to be doing during that time.

As an ancillary note, blocking out time for you and your mental health is also important. Exercise important for keeping you sane? Need time for activities with friends or you go ballistic? Whatever you need to keep you sane and happy, make sure you schedule time for it. Because when you’re insane, your work is never as good as you think it is. Trust me on this.