Sponsored Ad:
See the top rated post in this thread. Click here

Page 3 of 3 FirstFirst 123
Results 21 to 25 of 25

Thread: advice for RA position

  1. #21
    Trying to make mom and pop proud
    Join Date
    Apr 2017
    Posts
    8
    Rep Power
    2


    3 out of 3 members found this post helpful. Good post? Yes | No

    Re: advice for RA position

    Sponsored Ad:
    I am clearly late to the party here but thought I can give colour to the above discussion, involving Kaysa and others (Disclaimer: I was an RA at a highly ranked institution, but was never responsible for evaluating applicants.)

    I think many RA applicants are unaware that RA positions can vary greatly in the way they are managed. Let me give two cases somewhat inspired by reality:

    1) A junior faculty has obtained access to a "rich" administrative/commercial dataset no one else is using, or has scoped out a way to collect large amounts of data stored on webpages or PDFs. These data have the potential to kickstart a series of papers. With some grant support, the professor hires a RA who will be focused on cleaning, formatting and documenting this dataset.

    2) A junior faculty has a good pipeline of papers making their way through the review process. The professor needs to revise the papers for submission and hires an RA to speed up the process. The RA's work will involve a mix of augmenting the professor's Stata code and running new analyses suggested by referees (say, run a simulation in MATLAB).

    The ideal hedge against the uncertainty is to be a CS whiz and have expertise in all the programming languages used in econ - Stata, R, Matlab, Python, ARCGIS, Fortran etc - but this is not realistic for most candidates, even ones at top schools.

    Now my hunch is that most RA jobs are more like case 1) than case 2). Say you start from scratch. Knowing the above, first you learn Stata, the lingua franca of empirical micro today. Then you invest in languages rising in popularity - Python, GIS, some relational databases. If you still have time, pick up Matlab and R.

    Most importantly, if you say you know a language you need to show you can use it as part of a research pipeline. This is where cold-calling professors for some data work may land you an opportunity - or take a class with a data analysis project at the end.

    ...with all that said, talking about coding so much is looking at the tree and missing the forest. IMO, the largest constituency RA programs service are top-of-class US Liberal Arts College students who need an intermediary to match them to top PhDs. If you are not in that group, take note of the competition.

  2. #22
    An Urch Guru Pundit Swami Sage
    Join Date
    Jun 2012
    Posts
    1,734
    Rep Power
    13


    Good post? Yes | No

    Re: advice for RA position

    Quote Originally Posted by DisraeliShrugg View Post
    I am clearly late to the party here but thought I can give colour to the above discussion, involving Kaysa and others (Disclaimer: I was an RA at a highly ranked institution, but was never responsible for evaluating applicants.)

    I think many RA applicants are unaware that RA positions can vary greatly in the way they are managed. Let me give two cases somewhat inspired by reality:

    1) A junior faculty has obtained access to a "rich" administrative/commercial dataset no one else is using, or has scoped out a way to collect large amounts of data stored on webpages or PDFs. These data have the potential to kickstart a series of papers. With some grant support, the professor hires a RA who will be focused on cleaning, formatting and documenting this dataset.

    2) A junior faculty has a good pipeline of papers making their way through the review process. The professor needs to revise the papers for submission and hires an RA to speed up the process. The RA's work will involve a mix of augmenting the professor's Stata code and running new analyses suggested by referees (say, run a simulation in MATLAB).

    The ideal hedge against the uncertainty is to be a CS whiz and have expertise in all the programming languages used in econ - Stata, R, Matlab, Python, ARCGIS, Fortran etc - but this is not realistic for most candidates, even ones at top schools.

    Now my hunch is that most RA jobs are more like case 1) than case 2). Say you start from scratch. Knowing the above, first you learn Stata, the lingua franca of empirical micro today. Then you invest in languages rising in popularity - Python, GIS, some relational databases. If you still have time, pick up Matlab and R.

    Most importantly, if you say you know a language you need to show you can use it as part of a research pipeline. This is where cold-calling professors for some data work may land you an opportunity - or take a class with a data analysis project at the end.

    ...with all that said, talking about coding so much is looking at the tree and missing the forest. IMO, the largest constituency RA programs service are top-of-class US Liberal Arts College students who need an intermediary to match them to top PhDs. If you are not in that group, take note of the competition.
    That's a very helpful summary and you should stick around to help answer applicant questions related to RA work. Many of the frequent posters here (including me) did not have full-time RA experience before grad school.

  3. #23
    Eager!
    Join Date
    Mar 2017
    Posts
    46
    Rep Power
    2


    Good post? Yes | No

    Re: advice for RA position

    Quote Originally Posted by DisraeliShrugg View Post
    1) A junior faculty has obtained access to a "rich" administrative/commercial dataset no one else is using, or has scoped out a way to collect large amounts of data stored on webpages or PDFs. These data have the potential to kickstart a series of papers. With some grant support, the professor hires a RA who will be focused on cleaning, formatting and documenting this dataset.
    I am applying to an RA position in top 10 institution. In the second round, they provided me a data set for ticket price from StubHub with a tricky research question. This data set is unconventional, I don't think many people having access to these kinds of data.

  4. #24
    Eager!
    Join Date
    Dec 2017
    Posts
    85
    Rep Power
    1


    Good post? Yes | No

    Re: advice for RA position

    Quote Originally Posted by DisraeliShrugg View Post
    I am clearly late to the party here but thought I can give colour to the above discussion, involving Kaysa and others (Disclaimer: I was an RA at a highly ranked institution, but was never responsible for evaluating applicants.)
    Great post, DisraeliShrugg! As I mentioned previously in the thread, I am currently an RA at a top institution and have helped with screening applicants. Hopefully my experiences can be useful!

    Quote Originally Posted by DisraeliShrugg View Post
    I think many RA applicants are unaware that RA positions can vary greatly in the way they are managed. Let me give two cases somewhat inspired by reality:

    1) A junior faculty has obtained access to a "rich" administrative/commercial dataset no one else is using, or has scoped out a way to collect large amounts of data stored on webpages or PDFs. These data have the potential to kickstart a series of papers. With some grant support, the professor hires a RA who will be focused on cleaning, formatting and documenting this dataset.

    2) A junior faculty has a good pipeline of papers making their way through the review process. The professor needs to revise the papers for submission and hires an RA to speed up the process. The RA's work will involve a mix of augmenting the professor's Stata code and running new analyses suggested by referees (say, run a simulation in MATLAB).

    The ideal hedge against the uncertainty is to be a CS whiz and have expertise in all the programming languages used in econ - Stata, R, Matlab, Python, ARCGIS, Fortran etc - but this is not realistic for most candidates, even ones at top schools.
    I think these two are a fair characterization of what most RA work is (although I would add that you are not limited to a junior faculty being your manager). That said, I'm not so sure that people seeking RA positions should necessarily worry about which one of these two they focus most of their time on. Both options provide enough latitude to distinguish yourself and your work such that you can receive a great letter of recommendation, and learn a lot about the research and publication process. In option 1) there is a surprising amount of independence for how to both assemble the data and create the final sample that you can impress your principle investigator with your ability. Similarly, for option 2) RAs can certainly have an imprint on how to perform suggested analyses by referees.

    The nightmare scenario that you really want to avoid is a data entry job. To prevent this I think that you should ask the professor if you have been offered the job what they expect you to work on in the first several months of the position, and how they intend for you to do it. A year or two is too long to spend in a position where you have no capacity to impress your professor with your work. Also, ask if the professor has previous RAs with whom you can speak as they might be able to give a good picture of what the work was like.

    Quote Originally Posted by DisraeliShrugg View Post
    Now my hunch is that most RA jobs are more like case 1) than case 2). Say you start from scratch. Knowing the above, first you learn Stata, the lingua franca of empirical micro today. Then you invest in languages rising in popularity - Python, GIS, some relational databases. If you still have time, pick up Matlab and R.

    Most importantly, if you say you know a language you need to show you can use it as part of a research pipeline. This is where cold-calling professors for some data work may land you an opportunity - or take a class with a data analysis project at the end.
    I agree, I think it is unreasonable to expect a candidate to know every useful coding language coming in. When friends ask me where they should start I tell them to master one statistical programming language and develop competency in one object-oriented programming language. For the sort of applied micro research that I do the ideal combination would be STATA and Python, but other fields might have different preferences. You might be surprised, but just knowing two languages well will set you apart.

    Correctly signaling your knowledge of programming languages is important when it comes to getting your foot in the door. While most RA positions I know of have a programming task intended to test directly your knowledge, that task comes after an initial screening of your resume and cover letter. If you apply to RA positions make sure to provide details on where you learned these language in your cover letter, even if it was just from a course that you took. It is much more credible when an applicant writes "I learned practical programming in Python through a course in numerical applied mathematics, where we implemented several optimization methods including the simplex algorithm and stochastic gradient descent" rather than "I also know how to code in Python".

    Relatedly, sometimes applicants will write "I am currently learning how to program in Python". This can be much better stated through a short description of how you intend to learn. Say that you are taking a course at your college, or a MOOC, or briefly describe a data set that you want to scrape.

    Quote Originally Posted by DisraeliShrugg View Post
    ...with all that said, talking about coding so much is looking at the tree and missing the forest. IMO, the largest constituency RA programs service are top-of-class US Liberal Arts College students who need an intermediary to match them to top PhDs. If you are not in that group, take note of the competition.
    Despite my long discussion above of coding, I generally agree. The competition is fierce. Many of the applicants that I reviewed already seemed competitive for top 15 schools. I, however, wouldn't necessarily limit the constituency to just liberal arts college students. The best RA-ships also draw from elite research universities and students who switched to economics later in their undergraduate careers as well.

    I'm happy to answer any further questions people might have.

  5. #25
    Trying to make mom and pop proud
    Join Date
    Oct 2015
    Posts
    5
    Rep Power
    3


    Good post? Yes | No

    Re: advice for RA position

    I came from a 40-ish liberal arts college, currently working as an RA at a top 5 institution. My suggestion is to apply as many as you can, as there is so much unexpected randomness in the application process. I applied for about 50 positions last year, got through first round in 13 positions, and received 3 offers (I decided to drop from other applications after I landed on the current one). Previous research experience is not very important. None of the 3 jobs I got matches with my previous research fields in undergrad. Ironically, I didn't even get any response from those few positions that I directly had relevant research experience.

    Are you also interested in Fed or think tanks? My instinct is that positions at top institutions care more about GPA. Most of the job positions will ask you to do a designed data exercise. Some positions only asked me to show them some codes I wrote for my previous projects. How you perform on the task is probably the most important determinant. Also be sure to express your interests in doing research during the interview.

    A little more about my current job: I have two supervisors, one is a young assistant professor, one is a senior professor well-known in field. The tasks from the young professor are usually more direct and specific, while the senior professor usually only give general directions and is less concerned with the coding details (there are many other people in the same project btw). The trade-off is obvious: If you work with someone famous, you may not be able to get enough attention from him/her since your professor know so many students who are more talented than you are. If you work with a young faculty member, his/her name will be less noticeable to the admission committee.

Page 3 of 3 FirstFirst 123

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. advice for RA position
    By underg2121 in forum PhD in Economics
    Replies: 0
    Last Post: 12-23-2017, 01:28 AM
  2. Replies: 1
    Last Post: 08-08-2017, 06:34 PM
  3. Where to look for RA position?
    By Jayd in forum PhD in Economics
    Replies: 4
    Last Post: 04-16-2017, 11:17 PM
  4. RA position
    By acctguy in forum PhD in Business
    Replies: 2
    Last Post: 11-01-2010, 12:13 AM
  5. RA Position
    By tbroker in forum PhD in Economics
    Replies: 19
    Last Post: 04-18-2007, 03:01 PM

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •