I donā€™t think Iā€™m the expert on outsourcing research, but I do work a lot with research assistants and subcontractors, and I have answered this question so often that I may as well write a blog post about it and hopefully get comments on it.

How to find a good research assistant?

Here is where I search:

  • Your own PhD students: for tasks that require field-specific knowledge (e.g., Asset Pricing literature reviews, coding).
  • Finance PhD students at other institutions if none of our students is interested. Ask co-authors and friends for recommendations and talk to them when you give seminars.
  • Non-Finance PhD/Masters students: for tasks that require other specific knowledge (e.g., legal research) or do not require field-specific knowledge but intellect (e.g., news searches), or require access to your institutionā€™s data. Ask for recommendations, and post vacancies in scholarship networks. There you will typically find good students from unprivileged backgrounds, often without external funding or parental funding, who will be more committed and driven to deliver good work.
  • Volunteers/parent-funded (instead of self- or scholarship funded). Unfortunately many tasks that I outsource are repetitive and intrinsic motivation is mostly not enough to get the work done. Make sure that you give volunteers tasks from which they learn and that are intrinsically interesting.
  • Move on if the fit is not good. You are wasting your and their time that could be spent with a better supervisor.

I look for: evidence of motivation, prior experience in related tasks, especially as a research assistant, experience with data work.

I select: based on a small sample of work, typically a small project. I tend to work with the same research assistants over a long time on increasingly complex projects.

As with co-authors, research assistants have strengths and weaknesses and you may find them good at writing but not coding, or the other way around.

 

How do I find a good freelancer?

  • Iā€™ve had good experiences with upwork (www.upwork.com), but there are many other websites with freelancers.
  • Prior research experiences with universities is a great plus.
  • Ask them to answer specific questions in the application process to see if you will get along (e.g., ā€œcan you deal with brief instructions?ā€ I hired the freelancer that answered ā€œYesā€ instead of the 5 others that wrote an essay about their responsiveness).
  • Move on immediately if the fit is not good. You are wasting your and their time and they will end up with a bad review.
  • There are many tasks for which outsourcing can not only save time but produce better results: formatting of slides and tables, copy-writing, certain types of coding etc.

 

How do you give instructions effectively?

  • Understand the freelancer/the RAā€™s background, what level of instructions they require, how much independence they are used to, and what they would like to learn from your project (e.g., academic writing/research skills for PhD students; practical knowledge and leadership skills for Masters students). Follow up to make sure those learning goals are met.
  • Give initial instructions as comprehensive and detailed as possible. Do not assume that others know the details that you know and care for the ones you care for even if they are experienced. Adapt to individual relationship styles once you know someone well (and the other way around).
  • At the same time, give context so that they know which direction to go (e.g., show them the paragraph of the referee report you want to address). Encourage broader feedback and judgment.
  • Give examples.
  • Encourage expressing doubts and give explicit means to express doubts (e.g., dedicated column in spreadsheet for comments). Listen to their comments: strange comments are usually a door to something you did not know before.
  • Give explicit space to save underlying information (e.g., copy the source paragraph to the spreadsheet; save the article in a specific folder, under a name such that it will be easy to find).
  • Do not hesitate to ask for summaries in the format you need even if you think their task should be higher level ā€“ it is usually more efficient to make these while working (e.g., references in text format, coding diary). Be insistent about good file naming and documentation habits or otherwise you won’t find anything afterwards.
  • Start with small milestones and hard deadlines, observe your research assistant/ freelancerā€™s working style and give more freedom or more deadlines if needed. Encourage them to be open about other deadlines and plan around them.

 

How do you manage the communication going forward?

  • If you ask others to use any online accounts donā€™t forget to change back your password afterwards.
  • Isolate communication in one work stream, by email forwarding rules or with slack.com. Encourage communication and answer timely, but do not let yourself get interrupted by all questions.
  • Revise instructions every day based on output and doubts until it converges.
  • If you have a team, assign different members to check each otherā€™s work before sending it to you. Supervision helps to spot potential mistakes also in own work and is a good experience to have for students.
  • Ensure your research assistant / freelancer is happy with the tasks, learns, and receives timely feedback. Think about what you would write in a recommendation letter and tell them while you work with them, while they can still change, not when it is too late.
  • As with anything in research: try different approaches, experiment until it fits.

What do you do to improve research outsourcing?