top of page
  • Writer's pictureGRR's blog

Disappearing in to the stats black hole

Daily walks in the winter wonderland have helped keep me a little sane these past months

Daily walks in the winter wonderland have helped keep me a little sane these past months

I was surprised to see that I haven’t published a blog post in a couple of months now. I blame it on the statistical black hole of madness. These past couple of months have been the hardest in my PhD so far for many reasons. Although my results will be coming soon, this post is about a PhD and how hard it is and how it makes you grow as a person whether you like it or not. This is a post for other grad students to know they are not alone, for future grad students to know what they might be in for, and for me to share a little of the inside grad-student life.

Life as a bear biologist

It’s not all hanging out in the woods and collecting data. There are months on end of looking at spreadsheets and trying to find patterns in those number. While grizzly bears are sleeping all cozy in their dens, us bear biologists are sifting through massive amounts of data day in and day out. It’s this time in front of the computer that helps us shape a better understanding of how bears use habitat while they are awake and wandering around.

Research shows that grizzly bears are sensitive to human use. But the reality of bear habitat is that there are a bunch of different kinds of human use in a bunch of different places. People can be riding bikes, hiking, riding horses, dirt-biking, or riding quads. People can also be driving big logging trucks and transporting oil and gas by road or by rail. All of this can be done in foothills, mountains, valley bottoms, mountain tops, and even grasslands. And all of that can happen at different rates and intensities in the day, week, season, or year. So the reality of grizzly bear research that examines impacts of human use is that there is a lot of things going on in space and time and accounting for all of that is near impossible. This means that research must be focused and that sometimes you have to just accept that you can’t know or predict everything… even if you want to really badly.

The statistical black hole

Knowing all the uncertainties and variables involved in grizzly bear habitat use, most biologists collect an abundance of data. GPS collars in particular generate way more data than any one biologists could ever need. The first black hole is when you lost focus. It’s easy to find something interesting with one test and then ask yourself how that would change if you looked at the data a little differently. For me, this mostly meant running a test and then wanting to know how that result would change with season, or grizzly bear age, or grizzly bear sex, or a combination of these factors. Sometimes I can catch myself and re-focus on the exact PhD related question I wanted to answer before I get too far down the rabbit hole, but sometimes weeks can go by before I realize that I’m not really doing something I need to do – I’m just doing something cool.

And that realization hurts deep inside because timelines are short and every day counts.

Mostly I see the mountains through my office window. I have the prettiest view from the statistical black hole!

Mostly I see the mountains through my office window. I have the prettiest view from the statistical black hole!

The second kind of statistical black hole is the one that is more demoralizing and disappointing. That happens when you really want to answer a question and you’re super focused on it, but you can’t find the exact right statistical test that will get you there. One of the things I’ve been trying to develop is a model of human use on trails in the parks; I want to use my real camera data to predict human use on trails that I didn’t sample. I spent a week running different tests, putting in or taking out different variables, just trying to find a reliable model. But nothing I did was really awesome. One of models did well enough to be included in an analysis looking at grizzly bear habitat use around trails, but I had to really simplify the model and split human use into just low and high categories (instead of being able to predict a more accurate number of how many people are on a trail). Eventually, I had to give up on my vision of an accurate human use model. I really fought with this model and it took every ounce I had to admit I lost.

The emotional roller coaster

Doing stats all day every day for months is hard not only on my brain but my emotions. In the past 4 months, I have ranged from feeling deflated that I did something wrong, to elated that I figured something out. I’ve yelled and screamed when things didn’t work (including throwing an old fashioned temper tantrum when my printer wouldn’t work one day) and I’ve run into bed and pulled the covers over my head while I sobbed uncontrollably. In the past two weeks I’ve quit my thesis about a dozen times. I’ve felt crazy smart when I found something new. I’ve felt incredibly stupid and WAY over my head when I can’t see why I get the results I get. I’ve laughed, cried, cheered, danced around my office, yelled at my partner, gone to bed early, stayed up way too late, drank too much wine, eaten too much chocolate, and even prayed for mercy (and I’m not a religious person). And I did all that with a herniated disc in my lower back and some of the most pain I’ve ever been in.

These past months have been the most challenging of my life.

I know a PhD isn’t supposed to be easy and I didn’t expect it to be. In some ways I actually get solace from that because I know that most other PhD students have been through exactly what I’m going through right now.

Letting go of perfection

This past weekend, I handed in my biggest and most stats-heavy chapter draft to my committee. It’s not perfect by any stretch but I can’t look at it anymore. One of the biggest challenges with a PhD is accepting “good enough”. When I get so wrapped up in every details of a project, it’s easy to keep perfecting it, to keep tweaking statistical analyses or graphs or add one more paper to the discussion or to just do this one more thing. But then a PhD takes a million years longer than it should or needs to. So at some point I had to accept “good enough” and let go of the idea that my PhD is going to be perfect and change the whole world in 200 pages. No PhD is every perfect, and neither is any biologist. And that is OK. That’s actually the whole point of learning is that you’re never really done because you can’t possibly know everything.

And this man... who always helps me stay just a little silly... through the tears, temper tantrums, and epiphanies!

And this man… who always helps me stay just a little silly… through the tears, temper tantrums, and epiphanies!

Never in it alone

My little critter bums... save me from myself all the time.

My little critter bums… save me from myself all the time.

One of the things that I’ve really come to appreciate in the last few months is my partner and friends and family. Not everyone can wade with you through the emotional madness that is a PhD. I’m lucky to have a patient partner who has cooked dinner for me when I was working late and consoled me when I was inconsolable and has never gotten mad when I lashed out in frustration. I’m blessed to have friends who will go day drinking with me and listen to me drone on about this stupid linear regression and how I hate it. I’m happy to have a mother who send me little emails of encouragement every morning like a note in my elementary lunchbox. I’m especially amazed to have a graduate committee who is encouraging, smart, and incredibly patient with me and kind. As I wade through this madness, I’ve come to appreciate kindness so much more.

These are the people (and animals) who keep me sane and grounded. Thank goodness!!!

These are the people (and animals) who keep me sane and grounded. Thank goodness!!!

8 views0 comments

Recent Posts

See All


bottom of page