Consider whether your research could have negative outcomes too

When you’re considering the value of working on a particular research problem, it may also be important to remember that research isn’t a monolithic force for good. Research has done a lot of good, but there are many examples of it doing a lot of harm as well. There is a long history of research being biased by the discriminatory beliefs and blindspots of its time, as well as being used to justify cruelty and oppression. Research has made warfare more deadly and has facilitated the development of intensive factory farming. Dual-use biotechnology research is intended to help humanity, but could, for example, cause a catastrophic pandemic in the event of a lab accident or if the technology was misused. While some researchers are trying to increase the chance that future artificial intelligence is safe for humanity, many more researchers are focused on making AI more powerful. 

While it isn’t realistic for researchers to foresee every way their research could be (mis)used, many researchers are trying to create frameworks for thinking about how research can do harm and how to avoid this. For example, if you’re interested in working on biosecurity or AI safety, you could explore concepts such as differential progress and information hazards. If you’re working on global health questions, it may be important to educate yourself about the concept of parachute science.

Reach out to others for feedback

At this point, we think it could be helpful to identify some experts who might be interested in talking about your collection of potential research questions, and reach out to them for feedback. Getting feedback might then help you to prioritise between questions, develop your methodology further or discard projects before investing too much effort in them. You could seek feedback via two strategies – firstly, by sending your brainstorming document to people asking for general comments, and secondly, by seeking out people who have specialist knowledge on specific questions you’re considering and asking for their feedback on those ideas.

Here are some ways of connecting with other researchers:

• Reach out to your existing connections
• Attend research conferences related to your field of interest and speak to relevant people there, e.g. 1-1s at EAGs could be a great place to reach out to people for feedback on research ideas on directions that we recommend
• Are there any local student and/ or reading groups in your area that focus on a research area that you are planning to work on? 
• Public Slack channels on your research area, e.g. List of EA Slack workspaces
  
  

When preparing to reach out to experts, keep these key points in mind:

• Give the expert relevant information about yourself (e.g. What is your background? What is the scope of the project you’re planning to work on?).
• Prepare a short agenda if they’ve agreed to call you and share it with them beforehand (although they might not have time to read it, many people appreciate having the option to consider topics of discussion in advance).
• Think about what your key uncertainties actually are and what kind of feedback you want from the expert. Would you like their overall reaction? Detailed comments? Feedback on the strengths and weaknesses of your research ideas? Specific suggestions to improve your ideas? Feedback on how you plan to use the outputs of your research project? 
• Consider having a brainstorming document ready to share with them.
• You might want to have a look at this and this for more information about how to prepare.

Exercise: creating a weighted-factor model

Choosing which factors you want to base your thesis decision on will help you to reflect on what is important to you. Once you’ve done the exercise above and gathered some feedback from other people about your ideas, think about how much weight you want to give each factor. Lastly, try to evaluate how the research questions you’re considering score on each factor. The outcome of this ranking can serve as guidance for deciding on a question and can help clarify your intuitions about which questions would be the best fit for your dissertation. Here is an example of a ranking of potential thesis questions using a weighted-factor model (WFM).

Refining your research question

Once you have settled on a research question, it is time to develop a well-scoped and viable research proposal. The purpose of the proposal is to identify a relevant research topic, explain the context of the research, define concrete goals, and propose a realistic work plan to achieve them. If you’ve already built a Theory of Change for your research question, we recommend adding detail at this stage to help you create a proposal. We also think it’s important to reach out to your supervisor or other relevant people in the field of your research interests to ask for feedback, as this will help you develop an appropriate methodology. 

Here are a few more tips that could help you with narrowing the scope of your research project or refining your research question:

• First, make sure you have a detailed model of the problem you are planning to address in your research. Who are the different actors involved? How can research help fill gaps in our current knowledge? What are the particularly neglected approaches and interventions for this problem?
• You will only be able to make a valuable research contribution if your project is focused. Break down goals into discrete tasks and summarise what you are actually going to do. We suggest you create a detailed plan for the first few months of your project, a less detailed but fully coherent plan for the first year, describe a direction you might take in the second year, and generate some ideas for the following years. This will help you understand how much work is involved in every step and evaluate what is feasible in the available time frame.
• Consider practical questions. What kind of facilities do you have? Do you meet the university requirements?
• Try to develop the smallest possible question that can be answered and that data can be collected on, then have conditional upgrades/sub-questions based on that. This can be ambitious, but each stage should be developed enough to not be overwhelming or too vague.
• Start with a research question that’s as simple as possible and that you’re confident will be successful. From there, you can slowly and incrementally work towards pursuing more complex research questions. 

Find the best possible supportive environment

There are many different types of PhD programmes available – from 3-year PhDs to which you apply with a very specific project idea, to 6-year PhD programmes in which the first years are dedicated to coursework. It is important to find the best environment for your studies, with crucial considerations including the university and its community, the supportiveness of the supervisor/lab and the availability of funding. This section has advice on these three points and aims to facilitate you reflecting on them.

How much does the reputation of the university where you study your PhD matter for an academic career?

This is a commonly asked question among students, and we have compiled a set of key insights based on conversations with 30 of our experts. 

• The general advice is that you should pick the most prestigious university or research hub that you can get into.
• The importance of your university’s reputation varies across regions, with the US and the UK placing more significance on it compared to Europe or Australia. For the US especially, you will likely get a much better education and teaching quality, as well as access to resources, from a more prestigious university.
• It is worth noting that high-quality research labs (and supervisors) can be found outside of big-name universities, as specific research hubs may exist elsewhere. 
• It is important to note that even researchers in the most prestigious universities can be poor supervisors. 
• Ideally, you’ll find a great supervisor at a highly reputable institution. However, if you have to decide, finding an excellent supervisor seems to be the superior consideration – see below. 
• The significance of the university’s reputation increases if your career aspirations involve influencing government, e.g. in policy roles.
• Outstanding research, impactful contributions to the field, and a strong professional network could potentially outweigh the importance of a university’s reputation. 

Find a standout advisor

We think it is very important to find someone who genuinely cares about your research question and who will make a lot of time to supervise you well. Further, your supervisor will influence how effective you are in your work and how much you enjoy the research, as they will be the primary person guiding you throughout your whole research process. Especially at the PhD level, your advisor’s network matters tremendously for how well- connected you are and what sorts of opportunities will be open to you. So, here are some green flags to look out for in a supervisor:

• They care about your research question (pitch your ideas to the supervisor and see how enthusiastic they are about the potential project).
• They have the skills to supervise your project (check if they have experience in the methodologies you want to use).
• They truly care about mentoring you well (ask questions about their mentoring style, get a feel for how you match as a person).
• Their previous and current students are satisfied with them as a supervisor (ideally the person has a good track record of supervising other students – arrange a meeting with at least one current or past student). 
• They are successful (e.g. based on their citation count and general prestige).
  
  

Sign up for access to our database of potential supervisors who work on the research directions we recommend. Here are more tips on finding the right person to supervise you. 

Financing your studies

Even if you get accepted to a programme, it does not automatically mean that you get funding as well. Here are some tips if you need to apply for funding independently:

Consider a wide range of funding sources, e.g. national scholarships, university scholarships, grants and foundations dedicated to specific causes, and excellence scholarships (e.g. Gates or Rhodes Scholarships). Here is our funding database which includes funding opportunities relevant to the research directions we recommend.

Reflection prompts

• Consider the university environment – Would you be happy to live in the city of the programme you are applying to for 3-6 years? Do some university environments offer a more stimulating environment than others? Are there other researchers with similar values or motivations to you in this research hub?
• Do you have any hard criteria for choosing the location for your PhD? For example, would you consider moving abroad for an exciting opportunity? 
• What do you already know about the application process? What uncertainties do you have and how can you go about resolving them?
  
  

We recommend that you make a list of the programmes that best fit your research interests and other factors that are important to you. Then, check the requirements and deadlines for each of them and write down the next steps you need to take to apply. We also recommend reaching out to people who have gone through the PhD programme(s) you are applying to to hear about their experiences.

Set out your next steps

Take a few minutes now to write down your next steps for applying to the programs you’re interested in.

It could be helpful to sign up for some accountability buddy schemes, ask friends to check on your progress, or to set yourself a hard deadline on some important next steps that you want to take. You could schedule some time in your calendar right now, or make a note in your to-do list about a task that you want to complete soon.

Reflection prompts:

• What information do you need to get right now?
• What are you uncertain about? 
• What is keeping you from advancing with your project and how could you concretely resolve this?
  
  

Examples for concrete next steps could be:

• Reach out to people for feedback on your brainstorming document
• Reach out to potential supervisors
• Apply to an EAG or other academic conference and make a list of people you want to speak to 
• Reach out to people who have gone through the program you are applying to
• Reach out to current PhD students about proposal examples
  
  

Here are some further resources that could be helpful for you:

• Tips on impactful research
• Resources and tools for research
• Looking after your mental health
• Our Effective Thesis Community
• Research internships and other opportunities
  
  

For more general career advice, there are some other organisations that could help you with 1:1 advising. We recommend the following:

• 80,000 hours offers one-time 1:1 advising calls about using your career to help solve one of the world’s most pressing problems. They can help you choose your focus, make connections, and find a fulfilling job to tackle important problems.
• Magnify Mentoring pairs mentees who are interested in pursuing high-impact careers with more experienced mentors for a series of one-on-one meetings.
• Probably Good is running 1:1 advising calls to brainstorm career paths, evaluate options, plan next steps, and to connect you with relevant people and opportunities. 
• Lastly, please leave us some feedback. Thank you!