,
36 tweets,
12 min read
Read on Twitter
Legit learning outcomes for my new MSc module on data management and visualisation, right? 
Another module development question: should I devote an entire 2 hour class to how to google Stack Overflow, or rush it and cover it in 1 hour?
"Computing for the Behavioral Scientist" github.com/jashubbard/psy… teaching python programming fundamentals, in jupyter notebooks, from @jashubbard
Readings in Applied Data Science github.com/hadley/stats33… from @hadleywickham
Best Practices for Using Google Sheets in Your Data Project
matthewlincoln.net/2018/03/26/bes… wise words from
@matthewdlincoln
matthewlincoln.net/2018/03/26/bes… wise words from
@matthewdlincoln
Here's @russpoldrack's thoughts on running an intro to computing (R) course russpoldrack.org/2018/03/to-cod…
"I consider computational literacy (including coding) to be essential for any student today (regardless of whether they are from sciences or the humanities)"
hell yes!
"I consider computational literacy (including coding) to be essential for any student today (regardless of whether they are from sciences or the humanities)"
hell yes!
Video: Data Sharing and Management Snafu in 3 Short Acts
this is equal parts odd and painfully representative
this is equal parts odd and painfully representative
Here, @LorneJCampbell has articulated the golden rule:
"Organize your data and code from the start with the assumption that you will need to share this information" lornecampbell.org/?p=116
"Organize your data and code from the start with the assumption that you will need to share this information" lornecampbell.org/?p=116
"our wishlist for what knowledge and skills we'd find in a well-prepared data scientist candidate coming from a masters program." github.com/brohrer/academ… from @_brohrer_
Chapter 3 of @djnavarro's Learning Statistics with R, "Getting started with R" …gstatisticswithr-bookdown.netlify.com/intror
From Psychologist to Data Scientist
neurotroph.de/2019/01/from-p…
- some resources and links to get you started
neurotroph.de/2019/01/from-p…
- some resources and links to get you started
Working on some motivational slides for using coding to do data analysis
#1: Gene name errors are widespread in the scientific literature genomebiology.biomedcentral.com/articles/10.11… Excel autoformat introduces widespread errors into the supplementary material of leading genomics journals
#1: Gene name errors are widespread in the scientific literature genomebiology.biomedcentral.com/articles/10.11… Excel autoformat introduces widespread errors into the supplementary material of leading genomics journals
Motivational Example #2: Graduate student found excel formula error was driving erroneous conclusion of celebrated economics paper theconversation.com/the-reinhart-r…
Original paper: Does High Public Debt Consistently Stifle Economic
Growth? A Critique of Reinhart and Rogoff
peri.umass.edu/fileadmin/pdf/… correct the excel errors and the conclusion *reverses*: high public debt *doesn't* stifle growth
Growth? A Critique of Reinhart and Rogoff
peri.umass.edu/fileadmin/pdf/… correct the excel errors and the conclusion *reverses*: high public debt *doesn't* stifle growth
#3: The prevalence of statistical reporting errors in psychology (1985–2013) link.springer.com/article/10.375… 1/2 of papers looked at had errors, "One in eight papers contained a grossly inconsistent p-value that may have affected the statistical conclusion"
Dare you check your last report: statcheck.io ?
More motivation for analysis-by-code. Not sure exactly what to put on this slide, but it's true
Next: assessment criteria for final project submission. What factors should good analysis code be judged on? Suggestions welcome
Exciting times: the first class of my graduate seminar on "Data Management and Visualisation" is today! Link for slides explaining the module motivation and overview: docs.google.com/presentation/d…
Adding 'recursion' to the module's list of fundamental concepts in coding. If anyone asks me to define it I'll just ask them to google it
Update on this: the module is going well, and I'm having enormous fun, helped by tremendously engaged students who (i think) have no trouble seeing the relevance of this material to more reliable and powerful research.
Yesterday we did fundamental programming concepts in Python. Hurrah!
This week I'm going to get them pair programming. If you don't have a partner to programme with, you can use a rubber duck en.wikipedia.org/wiki/Rubber_du…
Parsons Problems: teaching coding by providing the correct lines of code, but in a jumbled order. delivery.acm.org/10.1145/116000… allows focus on control logic, rather than syntax, formatting. Definitely will try this on this course
This sounds insanely relevant. Amazed I hadn't heard of it before
As is this, with respect to the megathread this has become
Spotted in class. Glad the key messages are getting across.
Parsons Problems for data management and visualisation with Python github.com/tomstafford/pa…
Adding a bit of the bigger perspective to today's R exercises sciencedirect.com/science/articl…
I have been brazenly telling people that "80% of machine learning is linear regression" for a while. Evidence: kdnuggets.com/2018/12/superv…
Top 10 Coding Mistakes Made by Data Scientists
kdnuggets.com/2019/04/top-10… this i̶s̶ ̶g̶o̶o̶d̶ confirms my prejudices
kdnuggets.com/2019/04/top-10… this i̶s̶ ̶g̶o̶o̶d̶ confirms my prejudices
Now: preparing slides for the final taught class of this module, on using the terminal. There was really only ever one title I was going to choose for this class...
Good final class today: taught them to use ssh for remote access, then plugged my laptop into the room sound system so each student could announce success when they'd logged in to my machine from theirs - lots of fun
