Getting comfortable with failure

Every time I fail at something, I think about how I really should write a post about it, but then the sting wears off, I forget about it, and I move on until the next time it happens and I think, I really should write a post about this… So here it is.

Getting comfortable with failure is one of the most important skills a scientist needs because we fail all the time. I fail to get grants. I fail to make an experiment work. I fail to have a paper accepted for publication. I fail to meet deadlines. I fail to be the perfect mentor for a student. Some of these are small failures and some feel really, really big. None of these failures necessarily mean I am a bad scientist or not doing something correctly. Failure is just a reality of the job.

When we run an experiment, the whole point of doing it is that we don’t know the outcome – if we did, we wouldn’t bother doing it in the first place. This can be very hard for research students to adjust to because all their earlier education presented science as a series of known facts. But in research, even the methods used to test a theory can be entirely new and therefore prone to failure. When you do finally get an experiment working, you can end up with negative results that contradict everything you thought you knew! Sometimes negative results provide an important piece of information, but often they are hard to interpret and therefore don’t tell you anything useful at all. As scientists, we are aware of these problems and work hard to design our experiments to minimize this risk, but it still happens and it can be frustrating and confusing.

Rejections

A sample of the rejection letters received by myself and colleagues.

When we submit grants and manuscripts, we receive a LOT of rejection letters. Sometimes you did everything right and you still don’t make the cut. I recently received comments from a grant application and one of the reviews was probably the best I have ever received. In the end, I wasn’t even shortlisted. I personally think these failures are the hardest because papers are an important measure of productivity (whether it should be or not is a different discussion entirely) and grants are what allow you to be keep working as a scientist. Failure at either one can feel like a serious setback to your career.

Everyone has different ways to deal with failure and I don’t think there is a ‘right’ or ‘wrong’ way to do it, so long as you are able to overcome that failure and move on. For what it is worth, there are two things that I think help me in particular.

The first is that whenever I fail, I remind myself that this failure is just a stepping stone to succeeding and a valuable learning experience. When a manuscript is rejected for publication, for example, the subsequent version is always better. When a grant application is rejected, again, the next grant application is better or has more pilot data included. That doesn’t mean it isn’t frustrating and that it takes more time and impacts your productivity level! But by the time you have failed there isn’t much you can do about that and the only option is to keep moving forward one step at a time.

I am aware that to some it might sound a bit trite for me to say that I handle failures by looking on the bright side. For me, this works for the smaller stuff, but I acknowledge that some failure is too big or difficult to be overcome with positive thinking.

The real trick for me is that I separate my self-esteem and ideas of self-worth from my productivity or status as a scientist. When I fail, it is not because I am inherently bad, lazy or stupid; I just happened to not get something correct that particular time and I need to try again. This can be hard to remember if you have a series of difficult failures happen all once or seem to build on one another. But I remind myself that even if I had a grant and paper rejected in the same week and all my experiments are a disaster, I have other things that I do well in my life and I have hobbies/interests/family/friends where I am succeeding. I have also built a support network of friends, family and colleagues. This keeps me grounded and helps me remember that failures are only temporary, and that we all fail until we don’t. When things get really difficult, I remind myself about work-life balance. I leave early for the day and spend time with my child, I go and see a matinee (ok I have only done this once but it was really fun and I plan on doing it again if things get really hard) or I go for a walk in the woods and remember why I love studying animal behaviour so much.

I think it would be helpful if scientists discussed their failures more openly. Often we present our results as a polished package with a straight story line that makes it all seem so easy. This is done to ensure the result are presented clearly and can be easily understood. But being honest about the hard work and challenges of science when we speak to colleagues and students and in our outreach, allows others to learn from our mistakes and to be emotionally prepared for when it happens to them and not feel as though they are the only one.

Today’s diary

People may wonder what a typical scientist does all day and I find that question really difficult to answer. I don’t have a ‘typical’ day as a scientist wears many hats and everyday tasks are hugely variable. Whether I am in the lab/office/teaching/field/conference has a big effect on what I will be doing on any given day.

Today is a typical day in the office…

7:20am   Woken by screaming toddler. While this may not be early by some standards, I am not a morning person.

8:00   Husband takes toddler to nursery. I quickly down a cup of coffee and leftover food from toddlers breakfast. I wonder how in the world he managed to get salt on the banana…

8:30   Ride my bike down to the Plant Sciences building to pick up undergraduate exams that need to be marked. Am told they need to be returned by the next day.

8:45   Ride my bike over to the Zoology Department building.

9:00   Have my second cup of coffee for the day. Write a list of what I need to accomplish today. Mark Exams. Organize having an old desktop delivered from one building to another, get IT to figure out how to log in, make a copy of a very important file. Follow along with a tutorial on using fancy image recognition software/learn how to be an engineer in 30 minutes or less. Make a composite image of my field site. Spoiler alert: In the end, almost none of that will be achieved.

9:15   Read and send a few emails. One is about finding a location for filming an outreach video. Another is about a new aquarium facility that is being built. I am the representative for all the users and am in charge of designing the different aquarium systems. I spend the next 45 minutes reading up on coral husbandry requirements.

10:00   I have a meeting at 11 and need to read some papers and get my head around some concepts before then. I have my third cup of coffee. Respond to contractor/project manager about requirements for the new aquarium systems.

11:00   Meeting with Marian Stamp Dawkins about a new experiment I want to run. We discuss the different methods other research groups have tried and how we can use a particular test to answer my question without over interpreting the results.

12:00    Switch to decaf tea because I think that might be slightly healthier than having more coffee.

2018-05-31 13.30.23.jpg

My desk at lunch complete with scattered notes, papers, lists and an orange that is at least two weeks old.

12:45   Meeting finished and I am super excited about the plan but realize I need a bit of money to try out my ideas. Quickly look up possible small grant opportunities. I have missed two deadlines but one grant doesn’t have any set deadlines. Perfect.

1:00  Receive an email saying nursery fees are going up due to inflation. Spend 5 minutes worrying about how my rent, food and childcare costs are going up every year but my salary isn’t.

1:05  Buy lunch and eat at my desk. Normally I do socialise and take a break.

1:30   Respond to emails about the new aquarium specs required for the next two and a half hours. There are six different rooms holding freshwater and marine fish as well as corals. I need to make sure everyone has an aquarium that suits the animal welfare and experimental requirements.  During this time, that old desktop I was waiting for arrived.

4:00   I realize I only have an hour left to mark those exams.

4:35   A thunderstorm hits. I quickly call my husband and ask for a drive home as I don’t feel like cycling home in this weather. Looks like I will be marking exams this evening as there is no way I can finish before my ride arrives.

5:00   Leave for the day and finally bin that old orange on the way out.

5:15   Pick up my toddler. Go home and start dinner. I fertilize my flowers and vegetables as it is best done when the soil is wet. I plant three watermelon seedlings.

7:00   Try to put my toddler to bed but the thunderstorm keeps him awake.

9:20   My child finally falls asleep. I consider marking those exams but decide I need to improve my work/life balance.

10:30   Sleep.

All done with fun in the sun

img_65742

Thank you to my awesome team and the Lizard Island Research Station staff

Well after almost a month, I am back from field work (and some holiday!) in Australia. I was there conducting pilot studies for a new research project I am developing. The aim of this trip was to test two methods I could use to track fish underwater using inexpensive cameras. Since manually analysing video takes sooooo long, I had to set up my cameras so that a computer could track an individual fish. I also wanted to use the same video to build a 3D model of the terrain the fish was swimming in so I could get an idea of the navigational routes the fish were taking. To get 3D information from video, we filmed using two cameras placed side by side. This creates a stereo system that acts the same way as your two eyes to give you depth information. If you overlap two images from the cameras, you can tell how far away an object is by how closely they overlap. Objects that are far away will appear in almost the same position, while ones that are closer will hardly overlap at all.

Method 1

A series of cameras on tripods were set up around the territory of a fish. In theory, as the fish travels within the area that the tripods cover, we should be able to follow them. The pro of this method are that we can put the cameras in and leave them for a few hours and we don’t scare the fish by being there. The cons, as we discovered, were that it was a bit tricky to set up at first, it requires a lot of cameras to cover a large area (I had 16 cameras) and the area you could cover was limited by the visibility conditions which could change while you were filming.

We ended up setting the cameras up on the beach first so we could find the best configuration to cover the most area while having overlap with the other cameras. Once we got the hang of setting it up, this method captured some great fish activity.

Here is what the footage looks like and a few of the visitors we captured:

Method 2

Using only two stereo cameras, a snorkeler followed an individual fish for about an hour. The potential drawback is that the observer can scare the fish and change its natural behaviour but I found that it only took about 5 minutes or less before the fish got used to being followed. The benefit was that we can see the fish regardless of the visibility conditions and we are less likely to lose sight of the fish if they go behind a rock. I also like this method a lot because you can cover more distance. The analysis is likely to be the tricky part but we have yet to start.

Here is what the footage looks like:

 

Here are a few photos of our time on the island.

This slideshow requires JavaScript.

 

So long Lizard Island and thanks for all the fish!

Field work!!!

In 9 days I will be leaving for Australia to conduct field work and I am very excited about the prospect of sun, having just survived a long and grey British winter. To the uninitiated, field work seems like a glorified holiday allowing scientists to visit glamourous destinations and jot down a few notes about the local flora and fauna. If only this were the case! In reality, field work is HARD work and often takes a lot of planning.

For my current field trip, planning started about 8 months ago. And I am only away for 2 weeks! I first had to decide what experiments I wanted to run and exactly how I was going to do it. At this stage, you need to decide where your work will be carried out, who will be going, the best time of year for the particular work, and what equipment will be needed. If the experiments are part of a new project, then the next step is to apply for funding. For this trip, I applied for 3 grants and received two of them.

My field work will take place at the Lizard Island Research Station on the Great Barrier Reef, Australia. This is a national park so permits are required before I can go. If I were collecting specimens or doing more than observing animals, I would also need a fisheries permit and ethical approval by a committee within Queensland (easier said than done when you are based in the UK).

lizard-island.png

Maps of Lizard Island, Queensland, Australia (from Google Maps).

Once that was organized, I then needed to purchase, make and test my field gear. The research station is in a remote location and there are no shops so food and equipment can only gets delivered to the island by boat every two weeks. Planning for this requires thinking of everything. If you so much as forget pencils, you will end up wasting days until some arrive. Or you can use your ingenuity and make a pencil out of random things you can find on the island but for every minute you spend trying to make things, is a minute of valuable and expensive work time you have lost. The better prepared you are, the more likely you are to collect enough usable data. To ensure I pack everything needed, I not only have to have a very clear idea of what sort of work I will be doing, but I also must anticipate everything that could go wrong and come up with (several!) back up plans. At the same time, I am limited by the amount of gear I can carry or fit on a plane, so it is a balancing act.

The aim of this trip is to film Picasso triggerfish swimming within their home range and describe their navigational behaviour. I want to know how far they travel, if they follow the same routes, and if any environmental factors cause them to change their behaviour. I will be using 16 underwater cameras to film the fish and I will write a program to track their movements. To make the computer program work, objects of a known size and appearance must be in view of the cameras. None of this equipment can be bought at a normal store and so must be custom designed and built.

Gear.png
In order to automate the analysis of video data using a computer, easy to see objects of a known size must be in view of the cameras. Different coloured balls (top left) will be placed on the tripods so we know where left and right is (top right). I will also be attaching black and white checkerboards to the tripod. I 3D printed special checkerboard holders for this (bottom left). There is quite a lot to be transported (bottom right)!

While all the preparation can take a lot of time, it is the collection of data that is really the hardest and riskiest part of field work. Working outside in wild environments is physically demanding work. Even on a warm and beautiful tropical island, swimming all day in the sun is tiring and leads to a lot of sunburns. Add in the unpredictability of weather, sharks, crocodiles, jellyfish, venomous starfish (my personal nemesis), etc., and a lot of things can go wrong. After a few weeks of field work, I am always covered in a million cuts and bruises! All these factors can also make it hard to collect data. Weather is a particularly common issue. You may travel all the way to your field site, only to have a dangerous storm hit, making work impossible. Or sometimes animals just don’t cooperate and you can’t even find them! I was once on a trip searching for nudibranchs (colourful sea slugs) and it took hours to find only a handful. The very next trip, when I wasn’t looking for nudibranchs, they were literally floating everywhere and hitting me in the face as I swam. You just never know what is going to happen in the field…

Despite all of the inherent difficulties of field work, it still presents an amazing, and scientifically important, opportunity to observe animals and collect data that is often more natural than what can be collected in the laboratory. And of course, it is a lot of fun!

Research with animals

For many, working with animals sounds like a dream job, however, any pet lover will tell you that while animals are amazing companions, they can be a lot of work. This is no different when you keep animals for research.

There are many types of research that require animals, ranging from bio medicine to psychology. The aim of my work is to understand the natural visual behaviour of fish and I keep tropical fish in my laboratory. In order to study natural behaviour, I have to ensure that my fish are kept as healthy, and importantly, as happy as possible. It can be difficult to keep a fish used to living on a busy coral reef from getting bored in an aquarium but my experiments themselves actually provide entertainment for the fish. In my experiments, fish earn food rewards whenever they do something correct. For this to work, the fish have to be willing to participate.

Before I can start using any animals, the first thing I need to do is receive permission from an ethics board. How this is organized depends on the country where the research takes place, but it generally involves a panel of people who review your plan of how you intend to use the animals and how they will be cared for between experiments. One of the panel members is usually a veterinarian who understands what animals require to maintain their health. The panel can grant permission, reject your application, or ask for improvements to your plan.

IMG_8914

Although the water is clean, sometimes tanks can get a build up of algae on the glass and require a thorough cleaning (top left). While it can take up a lot of time, there is something really satisfying about a tidy tank. Regular animal care is a good opportunity to watch your animals and recognize how they behave when they are happy and healthy.

Once you have permission and have set up your laboratory, comes the important and ongoing task of keeping your animals happy and healthy. Animal care can take up a tremendous amount of time and is one of the biggest surprises for students new to working with animals. No matter what deadlines or personal events you have coming up, taking care of your animals is always your number one priority. This means that if a water pump breaks on a Friday evening, a pipe burst at 3AM or a fish needs feeding on Christmas day, you have to be there. Over the years I have missed or been very late to many personal events including a friends hens night that I had helped organize. Students who want to work with animals must be prepared for the realities of  animal care and must be flexible about working hours.

Some groups will be lucky enough to have a full time technician take care of their animals, but this really depends on funding, the size of your lab group or the type of animals you work with. While this can ease your workload, for scientists that study behaviour, there are benefits to performing the daily care of your animals. By spending so much time with them, you know when your animals are healthy and you get a good idea of what they look like when they are happy and relaxed. If you notice this behaviour changes at all during experiments then you know to rethink the task you are asking the animals to do. It is for this reason that I always try to get students to take charge of animal care and to spend as much time as possible with their research subjects.

Getting crafty in research

One of the things I love about my field of research is the opportunity to build things. Like a child building a rocket ship out of toilet paper rolls, I often don’t have the perfect materials or tools for the job so I have to improvise. When you are in the field and have a strict deadline and no nearby shops, you discover that you can build almost anything with some cable ties and other peoples garbage. In my recent project I had access to a few more gadgets than usual and have come up with something a bit more sophisticated.

The goal was to set up an experiment where I would train fish to swim down a corridor and go through a series of sliding guillotine doors. I had to make sure some doors moved at the same time and I was worried that me opening and closing doors would disrupt the behaviour of the fish. What I needed were some automated sliding doors.

Finished product first…

The first step was to come up with the best way to move the doors up and down. Should I use a pulley system? A series of gears? My choices were limited by the actual placement of the motors which had to be above the experimental aquarium and by the distance I needed the doors to move. After a few hours of online research I decided on using a rack and pinion system attached to a servo motor.

Physically making the custom gears for a rack and pinion would be pretty time consuming but I have a 3D printer in the lab which allows me to design and print my own creations in a matter of hours. The gears took a few tries to get right as the size of the teeth and the wheel had to be just right. I ended up finding an online calculator for rack and pinion gears which was a lifesaver. I made 3D drawings using Autocad Inventor. If you are a student, you may be able to get your hands on a free educational license for this software. Servo motor

With the design for the gears sorted, I then needed a motor to turn the gears. I used the Parallax Servo motor (https://www.parallax.com/product/900-00008) as this type of motor allows you to move the motor forwards and backwards, resulting in the up and down movement of the doors. This motor is also fairly cheap and can be easily controlled with an Arduino.

The motor arms had to attach to the pinion. Luckily for me, there was a design on Github for attaching something to the arms of this particular motor which saved me a lot of time. A housing was also made to hold the servo in place.

The motors and gears were then mounted above the tanks to line up with each door.

2017-09-11 15.28.37

Motors and rack and pinion gears mounted above the sliding doors

The next step was to attach the gears to the sliding doors. A very large aquarium was used for this project and I designed walls with sliding guillotine doors to fit inside. All the interior doors and walls were made with PVC and glued together. While I often have to make these structures myself, I was lucky enough to have the Oxford University workshop make this particular setup. I used cut aluminum rods to connect the doors to the the 3D printed gears.

The next task was to get the doors moving at the push of a button. For this, I used an Arduino Uno to send the pulses required to rotate the servo motor. The power output on one of these boards is limited, which would be fine if you are only using it to run one servo motor, but five motors require their own power adapter. As I wanted the doors to move up and down, I bought an on-off-on switch from RS components. I also bought a box to hide all the messy wiring.

2017-09-11 15.28.37-4
Electronics box with control buttons

My control panel had 4 buttons and had to move the doors up and down. After some more digging online (thank you Google!), I found a way to program the Arduino to run multiple push buttons on one circuit. The circuit board was then made and wires were run from the box to the servos.

Finally the doors were ready to try…

Although a lot of mistakes were made along the way, this ended up being one of the few projects where everything worked on the first try!

How do fish make a decision when faced with conflicting signals?

My newest paper called ‘Fish use colour to learn compound visual stimuli‘ has been published in the journal Animal Behaviour.

The aim of this work was to understand how fish respond when two pieces of visual information that they previously learned is conflicting. For example, there is a psychology experiment called the Stroop Test in which observers are presented with the name of a colour printed in either the same colour or a different colour and subjects are asked to read the word aloud. You can try this yourself with the following words:

RED BLACK BLUE GREEN

When people are asked to read the words composed of incompatible colours (e.g. BLACKBLUE) there is little change in how quickly they can answer. However, when people are asked to say the colour when the word is incompatible, they take much longer to answer (Stroop 1953, Macloud 1991). This suggests that when two pieces of visual information are in conflict, we rely on some information more than others. In this case, people rely on the word information more than the colour information, at least in this particular experiment. Stroop suggested that this occurred because subjects have more practice with one task (i.e. reading words) than the other (i.e. identifying colours).

We wondered if fish also pay more attention to one type of information than another. To test this, we used Picasso triggerfish (Rhinecanthus aculeatus) and trained them to discriminate between two circles that had both pattern (stripes vs cross) and colour (blue vs yellow) information.

Newport et al 2017 figure

We then changed the circles so that the patterns and colours were switched and therefore in conflict…

Newport et al 2017 figure 2

What we found is that the fish trained to choose the blue stripes, chose the blue cross when we changed the circles. We saw a similar pattern when the fish were trained to yellow crosses as they chose yellow stripes when we tested them.

This showed that the fish made their decision based on the colour of the circle and not the pattern.

Experiments like this are important because they help us understand what fish can see and what they think about the things that they see. This tells us a lot about how fish brains work and how their brains are different from the human brain. In this case, we learned that, just like humans, the fish brain can process colour and pattern information separately. It was equally possible that the fish wouldn’t be able to make a decision between the new circles because they weren’t identical to what the fish had previously learned. So even though the fish brain is a lot smaller than the human one, and made up of different structures, in this case, the two brains solved a problem the same way.

This experiment also shows us how important colour is to this species of fish. Picasso triggerfish live on extremely colourful coral reefs  and our experiments show that the fish likely use this colour information to tell objects apart.

Archerfish can discriminate human faces!

In June 2016 I published a paper showing that archerfish are able to discriminate between two pictures of human faces. They were then able to discriminate the learned face from a series of faces they had not seen before. This video summarizes the experiment:

This is a very exciting finding as human facial recognition is a really complicated task even for computers. Facial recognition is difficult because all faces have the same general components (e.g. two eyes, a nose and a mouth) so you have to discriminate them based on subtle cues (e.g. the shape of the eye, the distance from the nose to the mouth). Considering how similar features can be amongst family members, these differences can be extremely subtle. The features themselves are also changing constantly when we make different expressions. In addition, the orientation of the face and lighting can make a huge difference in what sort of information is available to the observer.

In my experiment, the fish were only presented with a frontal image of a human face under standard lighting conditions. This is a long way from full human facial recognition, however, it shows that fish are able to discriminate very complicated visual images and I will be continuing to test the fish under more difficult conditions in future experiments.

This paper is open access, so you can read it here for free. There has also been a lot of media coverage including The Washington Post, BBC, Livescience, Motherboard, The Guardian, Wired, CBC, Vox, The Huffington Post, Business Insider UK, Phys.org, Big Think, Mental Floss, IFL Science, CNN, Smithsonian, Gizmodo, and many more all around the world (it was apparently picked up by 175 news outlets). It even made it into the BBC’s ‘100 things we didn’t know last year list‘ at #15.

You can also see interviews I gave to Reuters, Quirks and Quarks and Deep Look on KQED.

The 3 second memory

Now that the fish have settled in, it is time to start putting them to work. In order to understand more about how the brains of fish work, I study their behaviour. While some researchers studying animal behaviour observe animals acting naturally in the wild, I run experiments in the lab designed to determine if fish can perform specific tasks.Whether or not they can perform a specific tasks can tell me important information about how their brains work.

A really simple example would be if we asked the question: do fish see colour? To answer this, we can present the fish with two identical objects, except that one is blue and the other is red. We can train the fish to select one of the objects, either the red or blue. If they can learn to select a particular object then we know the fish are able to see two colours. Now there is a lot more to testing colour vision, but this is the general idea of how I use behavioural experiments to answer complex questions.

I mentioned that I train fish… When I say this, most people give me a blank look and then ask: ‘but don’t fish have a 3 second memory?’ So before I explain how this is done, I need to pause and just clear up a little myth about fish.

FISH DO NOT HAVE A 3 SECOND MEMORY!!!

Ok now that we have that out of the way… To train fish, I use operant conditioning. Operant conditioning means that I use a positive reward to reinforce a behaviour that the fish would naturally do. In this case, the fish will naturally bite at things that are put in their tank. When they bite at the item I want, I reward the fish with a piece of food. This form of training only works if the fish are interested in performing the task. Luckily for me, triggerfish are curious by nature and love to bite at whatever I put in their tanks. And of course they love food!

At the beginning of this video, a fish is behind a white wall with a sliding door. When I open the door, the fish swims through. There they can see another white board with four circles on it. These circles are printed pieces of paper that are laminated. I stick them to the board with velcro dots.

You can then see the fish take a bite at the darkest one. This is the correct response from the fish, and I therefore feed the fish using forceps (because they bite!). My favorite bit is how the fish then swims back through the door to wait to start the process all over again.

I love this door that I had made as it provides a clear indicator to the fish of when a trial begins, and it protects me from being bitten when I put the stimuli board in the tank. However sometimes the fish get impatient with me and start spitting water at me out of the tank. Such demanding little fishies!

The triggers have arrived!

2015-12-10 10.23.08.jpg

It has taken a couple of months to get my new aquarium system ready for experiments and I am so excited that I finally have the fish I will be using for actual experiments.

For my research, I am using Picasso triggerfish (Rhinecanthus aculeatus) as my study species. These beautiful fish have so much personality. They are curious and expressive and will pretty much bite at anything. When working with them, you have to be careful not to leave your hand in their tank or you will get a surprising nip taken out of you!

You can see in my video below that they will attack things that come too close to the glass of their tanks as well (in this case it is my finger).

This species is found on the Great Barrier Reef in Australia and on coral reefs around the world. They are omnivorous and territorial. They spend much of their day swimming around their home region taking exploratory bites out of pretty much everything

photo-01-10-2016-7-51-58-pm   photo-01-10-2016-7-52-04-pm

In their tanks at the lab, the fish seem to enjoy rearranging the gravel by picking it up with their mouths and dumping it elsewhere.