Why do hummingbirds like red?

Earlier this year, we published a paper about the choices birds make at different coloured bird feeders[1]. Last week, I was contacted by a TV researcher with a question about hummingbirds. He’d seen a video clip of a hummingbird trying to feed from a man’s orange baseball cap, and wanted an explanation as to why this might happen. The obvious answer is that hummingbirds like red (and reddish colours), but do we know why? Asking a behavioural ecologist “why?”? I had to know.

First a disclaimer: I am not claiming to be an expert in hummingbird behaviour, pollination ecology, or floral trait evolution. I don’t even live anywhere where there are hummingbirds. As part of the bird feeder project, I briefly looked into the literature on choices at supplementary feeders, which happens to be entirely (until we published out paper) about hummingbirds. The first thing that jumped out from that brief survey of the literature was that the assumption that hummingbirds like red is actually wrong.

Why, then, is the hummingbird attracted to the hat?

Generally speaking, hummingbirds feed from red flowers, and flowers that are pollinated by hummingbirds tend to be red. So, this leads naturally to the assumption that hummingbirds like red. This is probably bolstered by the fact that if you wanted to buy a hummingbird feeder, you’d probably have to look quite hard to find one that wasn’t red.


Mexican violet-ear (Colibri thalassinus) at a flower. Image: Mdf (CC BY-SA 3.0)

Instead of asking why hummingbirds like red, we can turn the question around and ask instead why hummingbird-pollinated flowers tend to be red (although they are not exclusively red[2]). There are actually two competing hypotheses as to why this is the case[2,3,4,5]. The first is that hummingbirds have an innate preference for red, and find it easy to locate red flowers (birds have excellent colour vision).  The second is that the red colour is less attractive to other potential pollinators, such as bees, who find it more difficult to locate red flowers (bees do not have red receptors, and so are sometimes considered ‘colour blind’ to red, although studies have shown they can discriminate). By either mechanism, flowers that are pollinated by hummingbirds have generally evolved to be red, even though they might be quite distantly related, in a case of convergent evolution.

For the first hypothesis, there’s a body of experimental work investigating preferences for flower colour in closely related species of plants suggesting a preference for red and orange flowers over yellow ones[3]. But if we look at experimental work on hummingbird choices at artificial feeders, we don’t actually see this pattern – there is no particular preference for red over other colours – except when the hummingbirds are trained to feed from red feeders, or are captured from the wild on or close to red flowers[1,3]. Other than that, one study that seems to show that when artificial feeders are put in new locations, red ones seem to be investigated first by hummingbirds, but afterwards colour makes no difference[1,3]. What’s important are things like location and nectar quality. So, the support for an innate red preference in hummingbirds is actually quite weak, and the answer is not nearly as simple as “hummingbirds like red”.

foraging bees

Foraging bee. Image: PGloutnay CC0

The other hypothesis, that the flowers are red to deter bees, has a body of empirical support[6,7]. For example, bees find it harder to locate red flowers when they are against a green foliage background[4], especially when there is no UV component[2], prefer white over pink flowers when there is no difference in nectar reward[5], and avoid foraging on red flowers[7]. Overall, there seems to be quite a lot of support for “bee avoidance” rather than “hummingbird preference” as an explanation as to why hummingbird flowers are red. But why might flowers evolve to avoid being pollinated by bees? Aren’t bees what spring to mind when you hear the word “pollinator”? Because bees are potentially less satisfactory as pollinators: they tend to eat some of the pollen, and are more likely to result in self-fertilisation for the plant[5]. By avoiding pollination by bees, the plant can retain more of its nectar resources for its ‘preferred’ hummingbird pollinator, thereby attracting hummingbirds due to the greater nectar rewards that are associated with red flowers.

So back to the original question, if hummingbirds don’t really have a preference for red, why the attraction to the hat? Experience might play a role here here. If hummingbirds are attracted to resources that are of a familiar colour (and we know they can be trained to particular colours), associated with higher nectar rewards, and the flowers (and artificial feeders) they are feeding on tend to be red, and if red is also conspicuous in novel locations, then the hummingbird might well choose to explore the opportunities offered by a novel orange object in their environment. On arriving at the hat, the holes may bear enough of a similarity to the narrow corolla tubes (the central part) of hummingbird-pollinated flowers for the hummingbird to try feeding from them.


Feeding hummingbird. Image: GeorgeB2 CC0

[1] Rothery, L, Scott GS & Morrell LJ. (2017) Colour preferences of UK garden birds at supplementary seed feeders. PLoS ONE 12(2): e0172422 doi: 10.1371/journal.pone.0172422
[2] Lunau K, Papiorek S, Eltz T, Sazima M. (2011) Avoidance of achromatic colours by bees provides a private niche for hummingbirds. J Exp Biol. 214: 1607–1612. doi: 10. 1242/jeb.052688
[3] Handelman, C & Kohn, JR (2014) Hummingbird colour preference within a natural hybrid population of Mimulus aurantiacus (Phrymaceae). Plant Species Biology, 29: 65-72. doi: 10.1111/j.1442-1984.2012.00393.x
[4] Rivest SA, Austen EJ & Forrest JRK (2017) Foliage affects colour preference in bumblebees (Bombus impatiens): a test in a three-dimensional artificial environment. Evolutionary Ecology 31: 435–446 doi: 10.1007/s10682-017-9893-4
[5] Bergamo PJ, Rech AR, Brito VLG & Sazima M (2015) Flower colour and visitation rates of Costus arabicus support the ‘bee avoidance’ hypothesis for red-reflecting hummingbird-pollinated flowers. Functional Ecology 30: 710-720 doi: 10.1111/1365-2435.12537
[6] Schemeske DW & Bradshaw HD (1999) Pollinator preference and the evolution of floral traits in monkeyflowers (Mimulus). PNAS 96: 11910–11915 doi: 10.1073/pnas.96.21.11910
[7] Gegear, RJ, Burns, R & Swoboda-Bhattarai, KA (2017) “Hummingbird” floral traits interact synergistically to discourage visitation by bumble bee foragers. Ecology 98: 489-499 doi10.1002/ecy.1661


New research project: Parental care

This week marks the launch of our new, University-funded research cluster on Parental Care. About this time last year, the University put out a call to fund new research clusters, providing money for a combination of PhD students and postdocs, to allow groups of academics to develop their research into exciting areas. Isabella Capellini, James Gilbert and I put together a bid for 2 PhD students and a PDRA, and were successful! We advertised the jobs and recruited some excellent new people to join our emerging research cluster.

Isabella and James both have a track record in parental care research. Isabella has just finished supervising an excellent PhD on the evolution of male care in mammals (well done Hannah!), and James has done lots of work on parental care in insects and has a string of smashing papers. Me? I have a few bits that might be loosely termed “reproductive biology” – stickleback nests[1], guppy sexual behaviour[2,3] and some theory about mate guarding[4] and protandry[5]* from my PhD days. So parental care is really a new field of research for me.

So over the next couple of weeks, our two new PhD students, Stephanie McLean and Yannis Dimopoulous, and postdoc Dr Andrew Furness are joining Isabella, James, James’s current PhD student Alex Austin and I to create a ready made research cluster of people all interested in the same sorts of things. Stephanie will be working empirically on fish, Yannis will be doing comparative work on parental provisioning in insects, and Andrew will also be working comparatively, on parental care and life histories in fish and frogs.

banner copy

Anemone fish protecting its spawn (Silke Baron CC-BY 2.0), Bees (PollyDot CC0), Common rocket frog with tadpoles (Crisco 1492/Brian Gratwicke CC-BY 2.0), Ducklings (Capri23auto CC0)

*When migratory birds arrive back at their breeding grounds, males often arrive earlier than females. This is protandry, and we asked why it happens.

[1] Morrell, LJ, Hentley, WT, Wickens, VJ, Wickens, JB & Rodgers, GM. (2012) Artificial enhancement of an extended phenotype signal increases investment in courtship in three-spine sticklebacks. Animal Behaviour 84: 93-101 [link] [OA version]
[2] Chapman, BB, Morrell, LJ & Krause, J. (2009) Plasticity in male courtship behaviour as a function of light intensity in guppies. Behavioral Ecology and Sociobiology 63: 1757–1763 [link]
[3] Croft, DP, Morrell, LJ, Wade, AS, Piyapong, C, Ioannou, CC, Dyer, JRG, Chapman, BB, Wong, Y & Krause, J. (2006) Predation risk as a driving force for sexual segregation: a cross-population comparison. American Naturalist 167: 867-887 [link]
[4] Kokko, H & Morrell, LJ (2005) Mate guarding, male attractiveness and paternity under social monogamy. Behavioral Ecology 16: 724-731 [link]
[5] Kokko, H, Gunnarsson, TG, Morrell, LJ & Gill, JA. (2006) Why do female migratory birds arrive later than males? Journal of Animal Ecology 75: 1293-1303 [link]

Are we moving the goalposts?

“A slow sort of country!” said the Queen. “Now, here, you see, it takes all the running you can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!” [1]

One of the aims (or learning outcomes) of a university education is often that students should become ‘autonomous learners’, in that, broadly speaking, they should develop into people who actively pursue their own education rather than being taught exactly what they need to know*. But whether or not students do develop into autonomous learners depends, fundamentally, on whether autonomy is a fixed trait (as supported by some studies) or something that can be developed over time, or through particular experiences (as supported by others). On top of this, it is all very well for universities to set out to develop autonomous students, but do the students see themselves as becoming increasingly autonomous?

This is something we (me, Dom Henri and Graham Scott) wanted to explore [2]. Over the course of a couple of years, we distributed 825 questionnaires to our students, across all the years of a university degree. Some of the students got the questionnaire more than once, so we have data from the same students at different time points, as well as data from different degree levels**. The questionnaire was designed to measure how autonomous students perceive themselves to be, with questions like “I take responsibility for my learning experiences” and “I enjoy finding out about new topics on my own”, scored on a 1-5 “how strongly do you agree” scale. The scores are combined together to create an score of how autonomous each student perceives themselves to be.

What we found was (initially) a bit disappointing. There was no evidence that students thought they got more autonomous as they went through their degree programmes. Does this mean that autonomy is fixed, and students actually don’t get any better at learning for themselves as they move through their degree? Are we wasting effort trying to make them into independent knowledge-seeking-learning-loving graduates? Well, no. There are other things associated with autonomy that we know increase during degrees [3], so why do students not perceive themselves as becoming more autonomous? Is it, we wondered, because we are moving the goalposts?

Self-perception is often associated with the feedback we receive. So a student feels good about themselves if they achieve the grades they expect, and feel more confident when they think they are where they ‘ought to be’ [4,5]. So perception of autonomy might well be affected by things other than actual autonomy, but because we didn’t measure actual autonomy, we can’t say how the two are related. But at university, we expect autonomy to increase over time (it’s one of our stated goals!), so we give tasks that require increasing levels of autonomy, so the gap between where students think they are, and where they think they ought to be, could easily stay the same as they move through their degree programmes.

Because we’re all ecologists at heart, we thought this sounded very much akin to a Red Queen Effect [6], where organisms must continually evolve in order to stay in the same place (not go extinct), named after the Red Queen in Lewis Carroll’s Through the Looking Glass [1]. So because we expect students to become more autonomous, we are effectively moving the goalposts as students progress through their degree. We do this in lots of other, perhaps more concrete, ways too, like expecting more in the way of critical analysis in essays as students move through their degrees. So perhaps students aren’t going to perceive themselves as more autonomous if we continuously expect them to become more autonomous, and the gap remains the same.


Alice & Red Queen

John Tenniel. Through the Looking Glass by Lewis Carroll. 1871. Wikimedia Commons


Does it matter if students don’t think they are as autonomous as we perhaps want them to be? Well, I think it might. Autonomy is a learning outcome because it’s something employers are looking for. When applying for a job, students need to be able to express various things that might be associated with their autonomy. If they don’t think they are autonomous, that’s might do them a disservice in their applications. That would probably be OK, if there weren’t gender differences in perception of autonomy. Females routinely (in our study and others) report themselves as being less autonomous, less capable, than male students, when there is no reason to expect that they are, and so they could be placed at a disadvantage in the job market. If you don’t see yourself as autonomous and capable, you’re not likely to say you are in an application. And that definitely matters.

Which means that we should do something about it.


*There are lots of different definitions of what constitutes an autonomous learner, but generally they include things like taking ownership of learning, confidence in their own ability, engagement with student-led learning, and ‘deep’ rather than ‘surface’ learning.

**Dear students. Thank you so much for agreeing to fill in these questionnaires so often. It’s really helpful and we’re doing it because we want to understand how you think and feel so that we can make things better for our students in the future.


[1] Carroll, Lewis (1871) Through the Looking-Glass and What Alice Found There, Chapter 2
[2] Henri, DC, Morrell LJ & Scott, GW (2017) Student perceptions of their autonomy at University. Higher Education [open access]
[3] Thomas, L, Hockings, C, Ottaway, J and Jones, R, (2015). Independent learning: Students’ perspectives and experiences. Higher Education Academy
[4] Ainscough, L, Foulis, E, Colthorpe, K, Zimbardi, K, Robertson-Dean, M, Chunduri, P and Lluka, L (2016) Changes in Biology Self-Efficacy during a First-Year University Course. CBE-Life Sciences Education, 15(2): ar19
[5] Fazey, DM and Fazey, JA (2001) The potential for autonomy in learning: perceptions of competence, motivation and locus of control in first-year undergraduate students. Studies in Higher Education 26 (3): 345-361
[6] Van Valen, Leigh (1973). A new evolutionary law. Evolutionary Theory1: 1–30

Some thoughts on PhD interviews

This post follows on from the one on tips for making PhD applications.

I once gave a talk to a bunch of careers advisors as part of an AGCAS event about academic careers, and it made me think about how to articulate what I am actually looking for in a PhD student. I found a good quote here, which sums up what I think: I am looking for someone, basically, who can:

  • Do the work
  • Write about the work
  • Talk about the work

If you have written a good application, and got through shortlisting to the interview stage, that is where you have to convince the interview panel that you can do these things, bearing in mind that there will likely be other people who have also been convincing enough on paper to be shortlisted. There are several ways that PhD interviews are carried out, but it seems to me that things have moved on from when I had my PhD interview (two separate chats with the two supervisors) and have become more formal, more like a ‘real’ job interview. In recent years in Hull we have used two approaches:

  1. A panel interview with the supervisors, plus a couple of more ‘important’ people (in our recent interviews, we had the Research Director for the School and the Associate Dean for Research for the Faculty). This seems more common when the funding for the project has been secured through internal or external competition.
  2. A panel interview with some important people not associated with the project, plus a less formal interview/chat with the supervisors. This tends to occur when the project is ‘competition funded’, meaning that there is funding for a number of studentships, but more projects have been advertised, and the University are looking to recruit the best students across all projects. The supervisor does get some input, mind.

You should try to find out the format the interview will take, if that isn’t clear in your invitation to interview. But regardless of the format, the panel are really looking to establish the same thing: will you be a great PhD student who is motivated and excited by the work, and who will complete in a reasonable time and, hopefully, produce some good outputs?

When you are preparing for an interview, perhaps “do the work, write about the work, talk about the work” isn’t the clearest guidance. There are various lists available online as to what skills/qualities a PhD student needs, but when I was putting together my talk for the AGCAS lot, I found this one from the University of Reading helpful in articulating my thoughts (it’s on page 3):

  • Ability to grasp concepts and reason analytically
  • Motivation and perseverance in achieving objectives
  • Capacity for independent thought
  • Organisational skills
  • Independence as a learner
  • Self-confidence
  • Enthusiasm/passion for topic
  • Nature and extent of any relevant previous experience

Think about how you can demonstrate these in preparation for the interview. You probably won’t get asked them specifically as questions but you can probably detect them in the questions that you do get asked. Remember a PhD interview is exactly like any other job interview – you are competing with other candidates you know nothing about, yet have to convince the panel that you are better than they are.

Interview questions

These are some of the questions that we asked in the latest round of interviews, but I think they are probably quite common:

Why do you want to do a PhD?
This shouldn’t come as a surprise, but sometimes it does seem to. Here we are looking to establish whether a PhD just seems like the next step, because you are quite good academically and not sure what else to do, or whether you have a true passion for, and understanding of, what PhD research involves (hint: it should be the latter).

Why do you want to do this PhD?
This is where you can demonstrate your enthusiam/passion for the topic. It’s where having done a little bit of reading can help you demonstrate a bit of knowledge about what it is about. I would not expect you to have done a full literature review, but a little bit of knowledge of why the question is interesting, both generally and to you, is good here.

How does your previous experience/academic background/etc prepare you for PhD study?
Here is your opportunity to highlight your skills and experiences that will make you a great PhD student. You might want to touch on your taught modules, you research project, your wider life. Think here about the types of skills we are looking to see highlighted and how you can evidence them. We actually sent this question to the candidates to prepare in advance, instead of asking them to do a presentation, although you may be asked to prepare a short presentation (stick to the time limit and practice practice practice if you are), or given nothing specific to prepare.

Could you tell us about a research project you have done?
You might have already touched on this in a question about skills (it depends on the questions) but this is where you talk about your UG or MSc research project. Pick one to start with and pick the one that is either most relevant to the project or the one that is easiest to give a concise and insightful summary of. Focus your answer on hypotheses and answers, rather than methods (see the relevant bit in my writing applications post). You should be able to give your ‘elevator pitch‘ here. Sound excited. Don’t assume knowledge of the topic on the part of the panel (especially the Important Person who is actually a Geographer/Engineer/Biomedical Scientist).

Something about what you enjoy about research
Be sensible here – saying you love everything is probably a lie. I am not that keen on doing experiments (it’s why I do models and get PhD students to do experiments). I like analysing data and I like writing papers. I don’t like getting reviewers’ comments or rejections, and I do like getting acceptance emails. Try to avoid saying you most enjoy something that isn’t part of the PhD you are interviewing for – we want you to enjoy what you are doing, not forever wishing you were doing something different.

Something about what you find most challenging about research
Again, be sensible. Research is challenging and there must be parts that you don’t like so much. If you are asked what you don’t like, use the word challenging in your answer. Challenges can be overcome. Dislikes might stay.

Something about overcoming problems (in research)
Something probably went wrong at some point in your project, so think about what it was and more importantly, what you did about it. Here is your opportunity to show that you can nicely balance (for example) solving a problem yourself and running to your supervisor for help.

Something about what skills you expect a PhD student to have (and whether you have them)
Here, we are trying to get to the bottom of whether you are prepared for PhD study generally, and that you have some idea of what doing a PhD actually involves. This can be an opportunity for you to highlight how great you are at science communication, or motivation, or that you have what the deputy head of my secondary school used to call “stickability”. Be a bit of a swot and have a look at the Vitae Researcher Development Framework. Think about whether there are skills that you would hope to develop during your PhD.

What would you do after a PhD?
Basically, do you know where a PhD goes and what you might do with one.

Do you have any questions for us?
You absolutely must have questions for the panel, it shows that you are engaged with the process. It is not just an opportunity for you to find out things you want to know although if there are things you want to know, you should definitely ask. Good questions might include things like opportunities for wider development, opportunities to attend conferences, what you can expect in terms of supervision, opportunities for career development things like teaching or being involved in outreach, how the project fits into the wider structure of the department. If the supervisor is there, you can ask more technical questions about the project, such as how they see it developing, what opportunities there are for following you own ideas, whether they see publication as a priority. Don’t ask questions about working hours or holidays. Save those until after the offer has been made.

Important things

Turn up on time. If you are worried about what to wear, my advice is smart-ish, but not like you work in a bank. Equally, not your fieldwork hoodie and jeans. Don’t go out and buy a suit and don’t feel you need to wear the one your Gran bought you for your Uncle’s wedding.

Focus your answers towards the project that you are interviewing for. You may (if you are lucky or write outstanding applications) have interviews for more than one project, so think about the project you are interviewing for at that time and make sure you focus on that particular one, highlighting relevant skills and experiences.

Keep your answers concise, yet give them depth. Take a moment to think about what you want to say and how you might structure the answer. Give clear and concrete examples to evidence what you are saying.

Be enthusiastic. Nerd out a little bit about the project or science in general.

After the interview

You would normally, I think, be offered the opportunity for a tour of the place with a current PhD student, which can be helpful in asking the important questions you couldn’t ask in the interview, like how much the beer costs and where the nearest coffee is to the lab. For me, this isn’t part of the formal interview, and I do make this clear to candidates, but be aware that the supervisor might seek feedback from their student, although they should tell you if they will.

Remember, the interview is also your opportunity to find out whether this PhD/place is really for you. Hate the place? Think the supervisor is creepy? Get really bad vibes from the current PhD students? Think carefully about whether you can manage to spend 3+ years there. Hard, I know.

My tips for PhD applications

We recently advertised for two PhD students to join a new research cluster on the theme of parental care, and have now shortlisted, interviewed and offered the positions, and having recently read a bunch of applications, I thought I would share some tips about applying for PhD positions, based only on my own experience (tips for an interview coming in a later post).

PhD positions are competitive, often highly so, particularly if they involve a very popular topic, charismatic megafauna or fieldwork in exciting locations. They can attract some excellent candidates from top notch institutions, often with first class undergraduate degrees and a masters (either underway or already completed). The advert may stipulate something like “must have at least a 2.1 in a relevant subject”, but honestly, that’s unlikely to be enough. If you are a current undergraduate student, you really need to sell yourself in your application to be shortlisted when you are competing with those who have already demonstrated their academic credentials by finishing their degree, and have more experience that you because they are doing or have done a masters. Even if you already have a first and a distinction, that won’t automatically get you shortlisted. So what do you need to do?


  1. Read the advert* carefully and decide whether it sounds like something you might fancy doing.
  2. Make an informal enquiry to the project supervisor. Ask if they can provide additional information about what the project might involve. Even better, think of some more specific questions about the project, show a bit of knowledge of the broad subject area. A bit more information can help you shape your application. If the advert suggests that you include a CV with your informal enquiry, then do so (two pages is fine).
  3. Identify the skills and experience that the project requires. Are particular skills mentioned? Are there other skills that might be relevant but aren’t explicitly mentioned, such a field skills for a fieldwork based project, or particular lab skills for a lab based project. Make some notes about how your skills and experiences match what the project is looking for. Have a look at the work coming out of the supervisors group, look at their website and find out a bit more about what they do generally. Read a couple of papers that might be relevant to the project you are applying for. Get a feel for how they do their research (what methods, what analytical tools) and how you would fit in to that. Don’t go all out at this stage though, save that for the interview.
  4. Think about why you want to do a PhD, and why you want to do this particular project. Many of the applications we looked at lacked a good description of why the project appealed. All PhDs are different and we want to know what motivates you for this particular one.
  5. Write your application. What you do with the information you have collected depends on the application procedure. We asked for a CV and a statement outlining this stuff, so my advice is based on that. It’s likely that the same information will be needed regardless of the application format.

Your CV

Your CV need to be clear and fairly academically focused. A CV for a PhD should probably different from one for a job outside academia. We are interested in your academic skills and research experience, less so in your part time bar work. Equally, tailor your CV for every application you make – which bits do you want to highlight most for this particular project? Things that should definitely be on there include:

  • Your degrees (these should come first), where they were and the years you studied them plus your overall or predicted degree classification
  • Perhaps include some relevant modules you have taken, which year they were in (first, second, third) and possibly the mark you got for those modules (although you will most likely also need to provide a transcript which has this information
  • Your research project from each degree – the title, the mark you got and a brief description of the research skills involved (perhaps in a “research skills” section)
  • Some information about other skills that you gained during your degree (“relevant skills” – think about research skills, analytical skills, communication skills, organisational skills)
  • Qualifications from school/college
  • Relevant work experience. If you have carried out any vaguely relevant volunteer or paid work, give a short description here. I might be advertising a behavioural ecology project on fish, but if you have volunteered in conservation, or science communication, or teaching in a school, that’s great, it demonstrates to me a broader motivation to work in something vaguely ecological/scientific/academic, and hey, I did some of that that stuff too before I became a PhD student. If you haven’t got anything ecological, don’t worry, but think about what your other work has given you that would make you an excellent PhD student. Independence? Rising to challenges? If you are going to mention a job in your statement (later), make sure the details of it (what, where, who, dates) are in your CV.
  • Hobbies/activities/other qualifications. Yes, list these, particularly if they are relevant to the project (applying for bird project and have a ringing licence? Definitely mention it, but make sure it’s clear why you have it and how you got it). I personally don’t find a list of your hobbies particularly relevant – what you get up to in your spare time shouldn’t have any bearing on whether you would make a good PhD student or not. Who am I to say that someone who cycles 200km at the weekend would make a better or worse PhD student than someone who makes YouTube videos about Minecraft?
  • Include the contact details for two academic referees. Make sure they can make a comment about your suitability for PhD study. Ideally, they should have experience of you as a researcher (such as your project supervisor). Your tutor might not be the ideal person for the job, particularly if you barely see them and haven’t taken any modules they teach in your final year. It is better if they can talk about you, rather than the marks you got. Contact your referees in advance, ask if they are willing and able to provide a reference for you and give them details of what you are applying for and a copy of your application.

Your statement

This is where you explain to the supervisor why you would be an awesome PhD student, why you are perfect for this PhD and why this PhD is perfect for you. My top tip is NOT to start with an “Ever since I was a child, I have always been inspired by the natural world. From watching David Attenborough as a two year old…”. Why does this make you better for the PhD than someone who thought they wanted to be an accountant but saw the light and made the decision later? It doesn’t. Focus instead on the recent experiences that make you great for this, I need to be convinced that you can carry out research, not that you like watching nature documentaries.

The recent crop of applications were very varied in length, but I would aim for about a page, no more than two (single spaced, sensible sized font) unless something else is specified (like a word count). In this page you want to convince me that you can and want to do this particular PhD. Convince me that you are interested in the topic, but you don’t need to convince me that the topic is generally interesting (I know that, or I wouldn’t be looking for a PhD student to work on it). If you have already worked on something similar, great, if not, it doesn’t necessarily matter as I am very aware that the little niche I want you to work in might not be the same little niche that you fell into as an undergraduate.

Tell me about the research you have done, but focus on the outcome and skills, rather than the method. A good summary might be something like:

“My project focused on testing the hypothesis that behaviour X occurs more frequently under condition A than condition B. I found that generally, my hypothesis was supported, but larger males were more likely to carry out behaviour X than smaller males under condition B. During my project I gained a number of skills that are relevant to this PhD project. I was responsible for the husbandry and feeding of my animals, I designed and piloted my experiments, then adjusted my methods to improve the accuracy of my data collection. I gained skills in animal observation, method X, method Y and method Z. I analysed my data using linear models in R, and produced a project report for which I was awarded 86%.”

Do similar things for modules where you gained relevant skills, but don’t make sweeping statements about having skills that you did once in a practical in first year and never did again. Make sure you tailor your application to the project you are applying for: Molecular project? Highlight those skills. Behavioural observation project? Highlight that experience. However, you can and should certainly mention any other skills you have but be honest with yourself as to whether you have the skills, or you once did a practical that used them but can’t really remember. Don’t think all your skills have to come from your studies – think about ‘softer’ skills you have from work, volunteering, clubs and societies etc, particularly your communication or organisational skills.

If your study of the advert and any extra information you have been sent has identified particular skills that you need, make sure that you clearly state that you have those skills and where you got them. A statement that says “I have experience with R” is not nearly as convincing as “I took a module in Research Skills which included statistical analysis in R (GLM and GLMER), and carried out the analysis for my project in R, where I learned to write simple functions to sort my data”. There will likely be enough applicants that I don’t need to dig to find out what you can do, so make it obvious to me. If you don’t have a particular skills but are awesome for the project in other ways, then indicate to me that you could learn the skill: “Although I have not yet had he opportunity to learn R, I have experience of statistical analysis including mixed effects models in SPSS and have programmed using Python during my module on X”.


Get someone to look over your CV and statement. Preferably, this should be one of your lecturers or someone who has appointed PhD students in the past. Get some tips from them as to how to improve and what they would look for. Proof read several times. Make sure you have included everything required by the application (a completed form, a CV, a statement, your transcripts, your English language qualifications, whatever is required by the University you are applying to). Hit the submit button and cross your fingers.

Good luck!

*These are tips for applying for advertised PhDs. If you want to find a supervisor and convince them to help you write a funding application, that’s possible too, but it’s a bit of a different process.

Writing your results section

Following on from my post about writing a discussion, here are some tips on putting together a results section for your project/dissertation (in no particular order). These are the sorts of comments that I regularly find myself writing as feedback on drafts. Maybe next year I will remember to give this to all my students first

1. Your results should focus on your results, not on the analysis methods. It is better/clearer to have a “statistical methods” section at the end of your methods, so you can focus your results section on telling the story of your data. In this section, you need to make sure that you justify each decision that you have made for the analysis (why did you pick that particular test? Why did you divide your data into categories in that particular way?)

2. You should always (at a minimum) include the test statistic, sample size (N) or degrees of freedom (d.f.), and p-value for every statistical test. Just reporting the p-value is not enough. If you are using different analytical techniques, follow the standard reporting for those techniques.

3. Report these outputs (and any other figures, like means or standard deviations) to 3 decimal places tops. You didn’t measure things to 9 decimal place accuracy, so don’t report with that level of accuracy either. If you measured body size to the nearest mm, you could report the mean body size to one decimal place, for example.

4. Always refer to your figures in the text, next to the analysis they are associated with. You can simply put the figure number in the parentheses after the stats, you don’t need a whole sentence that says “Figure 2 shows this relationship”. Each figure should have a clear legend that allows it to be understood without reference to the text. To stop stuff moving around when you get me to comment on it, paste your figures “in line with text” and write the legend underneath, not in a text box. You can pretty it up at the end before you submit. If you have multi-panel figures, label them a, b, c, d and refer to them as figure 1a, figure 1b rather than “top left” and “bottom right”.

5. You don’t need to explain in general what a particular test does, just why you used it: Say  “To assess whether there were differences between categories, an ANOVA was used” rather than “ANOVAs test for differences between categories, so I used one to test whether…”

6. Ditto for figures. No need to explain that a scatter plot shows the data points. There might be an exception here if you have used a particularly complicated plotting technique.

7. Make sure you know the difference between relationships (between two variables) and differences (in your response variable between categories) and which one you have tested for.

8. Tell me what the analysis you have just reported actually means. “There is a statistically significant difference” isn’t terribly informative. “Men are significantly taller than women” tells me a lot more. Always give the direction of an effect (was it a positive or negative relationship, was A larger or smaller than B?). Follow your statement with the results from the analysis.

9. If your results are not significant, use “non-significant”, not “insignificant” and avoid making any comment that suggests a relationship or a difference that isn’t supported by a your analysis (even a non-significant one). That’s why we do the analysis, to avoid drawing conclusions from eyeballing the data.

10. Look back at the statistics courses you have taken during your degree and get guidance as to how to present your results from there. Stick closely to the format. If you’re a Hull student, this can be found in the 2nd year Skills Stats Handbook (you all got one). Another source of inspiration is the results section of published papers that use similar analyses to yours.

11. Don’t worry about your results section seeming short. If you have correctly reported all the analyses that you should have carried out (i.e. done everything sensible, given your data), based on your hypotheses/aims/research questions, then your results might not be enormous, but so long as they are written correctly, that’s OK.

The tighter and better written your results section, the easier it is for the reader to follow. Key though, is that you understand what you have done, why you have done it and what it means. This needs to come through in your writing.

Writing your discussion

Each year, at about this time, I get emails from my project students, asking questions about how to write up their projects, particularly the discussion. Each year, at about this time, I knock out several versions of an email explaining how I would go about writing a discussion. So this year, having knocked out the first version, I have decided it would be a good idea to save it for posterity, and for the next project student. So here we have (as sent to the student):

How I write a discussion

For your discussion, I think the key thing is to make sure you are actually discussing your results. This might sound silly, but so often I see discussions which are a literature review and don’t relate that literature to the findings of the project.

Start with a summary (one paragraph) of what you found in relation to your hypotheses. Don’t explain the hypotheses again, or review where the question came from (you did that in the introduction) but focus instead on what you found. Make sure you really highlight the interesting/key findings here. It’s not a repeat of the methods or results, so focus more on what it means (as in, don’t talk about significant differences between things, but talk about support for hypothesis A and no support for hypothesis B, making sure you mention briefly what they actually are).

Then, the way I think about it is to pick out each individual result you want to discuss and think about how you can devote a paragraph to that result. Think about whether the result fits with existing literature or contradicts it, and make a statement to that effect (“the finding that large fish ate more is consistent with previously published research”), then review that literature. Make sure you are doing some critical thinking/writing, so bring together examples that show the same thing and contrast them against examples that don’t, rather than just listing studies. Try to tie every paragraph to something about your results or your project generally.

I find a helpful way to approach this is to write subheadings (later deleted) for each paragraph, such as “fish size and appetite: big fish eat more” and “fish size and appetite: why do big fish eat more?”. I can then draft out each paragraph under the heading and if necessary, split or combine paragraphs later. I keep the subheadings in until I am happy with the overall discussion, then delete them.

For each paragraph, think about the PEE framework*. Point. Evidence. Explain. Start the paragraph with the point you want to make, give the evidence to support that point, the explain how the evidence supports the point (this part might be obvious). The key thing is to use plenty of literature here and think about what your results mean and how they fit in with what we (the scientific community) already know.

Think also about the limitations of your study and how it could have been better. There are two ways to do this. One is to have a ‘limitations’ paragraph (or two) where you say what you might have done differently and why. The other is to intersperse the limitations with the rest of the discussion, so for example if you wanted to mention that one limitation was that you only tested one group size, you could drop that in as an opportunity for further work, or something to be considered, at the end of a paragraph where you are discussing something related. If I am writing a paper, I prefer the latter approach, unless my limitations/opportunities don’t fit well with the other key things I want to discuss.

Try not to turn the limitations section/bits into an opportunity for self-flagelation. I have seen projects where there are several paragraphs about how sh*t the project** was and how nothing went right. Rarely do you need to do that and it is much better to focus your discussion on what you did find, rather than what you might have found had you had 15 years and millions of pounds to carry out the research. If you can, turn limitations into opportunities for future research, and remember to use references here to support your ideas.

At the end, you want to try to bring everything together into some sort of conclusion/wider context paragraph. Don’t do too much restating of what you have already said, but also don’t bring in too many new ideas here. The idea is to round everything off and end on a high note/take home message.

In terms of the word count***, I don’t think there is any particular expectation for the discussion. In my opinion, it is better to have a well-written, focused discussion that covers all the key areas, than to have a long rambling discussion with the sole aim of getting it up to some sort of arbitrary word count. As the handbook says “there is no benefit to excessive padding”.

More useful resources

I have started collating together some useful websites for projects and writing, have a look here. Let me know if you find it helpful or if you have come across other useful resources while doing your project.

*I first learnt about the PEE framework a few years ago from a colleague whose kids were doing it in school. It’s apparently part of Key Stage 3 English (KS3 is years 7-9 of school, or kids aged 11-14ish). It doesn’t work for every paragraph, but if you think about writing paragraphs using this sort of framework, you can tweak them later to flow better. When we first started mentioning it, our students faces were blank, but now some of them do acknowledge that they have heard of this.

**I’m not denying it, some of these projects were supervised by me. They weren’t sh*t, they just ended with non-significant results, but this often translates in a student’s mind to “this project didn’t work”. I wonder if we perhaps don’t given them many opportunities to do practical work that doesn’t succeed in achieving whatever we wanted it to achieve.

***Students always worry about the word count. I worried about the word count when I was a student too, but mostly in terms of fitting everything I wanted to say in to it, rather than including enough stuff to reach it. I have to remind myself (as all academics should on a regular basis) that I was not a typical student. Still a problem, only now it’s character counts and page limits.