Skip to content

Instantly share code, notes, and snippets.

@colaresi

colaresi/GraduatePolSighAdvice.md

Last active August 31, 2015 23:53
Show Gist options
  • Star 1 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save colaresi/e4f3283f7c5e9416d83f to your computer and use it in GitHub Desktop.
Save colaresi/e4f3283f7c5e9416d83f to your computer and use it in GitHub Desktop.
Download ZIP
Some Unsolicited Advice About Tools for Modern Political Science Research
Raw
GraduatePolSighAdvice.md

Some Tools for Modern Political Science

First Year Graduate School Edition

by Michael Colaresi

The Point

'All advice is wrong, some advice is useful'

A slight amendment of George Box' famous quote

This list is necessarily incomplete. Not just because everyone's research is like a snowflake and different (althought that is true) but because the point of learning research tools like those listed below is to train yourself to continue to pick up new things -- whatever is necessary to move your research forward -- throughout your career. Too many people view graduate school as the time they learn how to use the one screwdriver (usually one that is flat) that will carry them through a career. But that gets things backwards, the goal is not any specific tool, but to learn how to efficiently invest the time how to get your research done. So on that note, here is some advice as to where it might be useful to start (but not end).

Math

`Do not worry about your difficulties in Mathematics. I can assure you that mine are still greater.'

Albert Einstein

`Go down deep enough into anything and you will find mathematics.'

Dean Schlicter

Since we are in the business of teaching students how to engage with research, not be ignorant of it, understanding mathmatical notation and concepts is essential. A few places to start:

Code

`I do not fear computers. I fear the lack of them.'

Isaac Asimov

`Man is a slow, sloppy and brilliant thinker; the machine is fast, accurate and stupid.'

William M. Kelly

The ability to write code is increasingly important. From collecting and collating notes, collaborating with coauthors and combining work efficiently, to documenting a replicable workflow, analyzing data and presenting results, not to mention in writing your award winning book (breath), you will need to code. Readable code not only helps you avoid mistakes, but allows others to build on your work.

  • Basic Computing

  • Do you use:
    A. Mac: Then go here, h/t Jeff Gill.
    B. Windows: Switch to Mac or Linux (no not really, but I can't help you).
    C. Linux and/or Virtual Environments: Then you probably know more than me, so please email me your version of this document.

  • Shell Scripts/Command Line:
    The command line is a bit like the bridge of a ship. It is where the controls are and the decision are made. If you are not on the command line, what is?

    • Software Carpentry has video tutorials that are terrific, see this link.
    • As well as this useful book.

  • Writing (yes, writing is code)
    Learn TeX:

  • R
    R is useful for many things, and it is free. To get some motivation, check out the gReat graphics at ggplot2 and the even better puns at magrittr.

  • More seriously, one might begin here: An Introduction to R, By W.N.Venables, D.M. Smith and R Core Team.

  • Or here: The R Guide by W.J. Owen. 2012

  • Then dabble in the plethora of packages, such as MSBVAR, MGCV and many others.

  • I would also check out R for Machine Learning, by Allison Chang, at MIT, Open Courseware. Then learn what the different algorithms R.

  • Python
    Now for something completely different. Python is a tool that slices (instead of indexing) and dices. Actually, I have used it for scraping and working with text mostly. But it has provided several eureka moments.

  • I like The Practice of Computng Python by MSU’s own William F. Punch and Richard Enbody.

  • I am an Anaconda user and have found that to be a terrific low cost way to begin.

-Version control
Git seems to be terrific in my limited experience. MSU has a Gitlab, so code can be private as you work on things without paying a subscription fee.

  • A gentle introduction is here

Research Workflow

`Science is organized knowledge.'

Herbert Spencer

One of the hardest things to realize, until it is too late, is that spending time being organized and thorough now may save you months later. I learned this from Scott Long (well I should have learned it), who was one of the best teachers I ever had.

  • He wrote a terrific book on workflows for Stata users, see his website for the book.
  • More generally, I really like the approach in Matthew Gentzkow and Jesse Shapiro, 2014. “Code and Data for the Social Sciences: A Practitioner’s Guide”. available here

Bayesian Statistics

`Bayesian statistics is difficult in the sense that thinking is difficult.'

Donald Berry from J.E.H. Shaw

`Credo ut intelligam.'

Saint Anselm from J.E.H. Shaw

One word: Update.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment