Quincunx via Processing (Sorta)

For my ICM final, I forecasted a month ago, collision was definitely the path that tortured me…again and again. The end result was about 80% there, but definitely didn’t do what I wanted. Thwarted again by Processing…the story of my semester.

Sketch

The Canon and The Lady Tasting Tea

In taking this class, I really gained an appreciation for authors who can write about statistics and methodologies well. The gold standard is Huff’s How To Lie With Statistics and dead last in the heat was The Canon. I actually found that book dreadfully dull and couldn’t get anywhere before I gave it up. The faux-naivete of the writer/narrator was totally noxious to me and I couldn’t stand to continue.

Another book more intriguing (but also with its faults) was The Lady Tasting Tea. I found myself torn about the book. On one hand, it was pretty fascinating to find out how a lot of the statistical theories and methodologies evolved over time. But the book also devoted way too much time in my opinion to the personal lives of some of the towering figures like the Pearsons and Fisher. I thought Gosset was the more interesting of the four and would have loved to read more about him.

How To Lie With Statistics

One of the first readings for the class was the amusing How To Lie With Statistics. I definitely got the sense while reading it that it had to be the Freakonomics of its day, with the same irreverent tone I know from those book and the blog. How to lie with statistics is a fairly simple task. It mostly involves presenting facts in a misleading visual style so the reader will be drawn to the false supporting conclusion the author is going for.

The book was a very quick read and entertaining, even if some passages were definitely dated (i.e. the “good” salary of the Ivy-educated man which approached the low $10Ks).

Human Subjects Certification Exam

Part of our class assignments was to venture into the world of social sciences at NYU and take the Human Subjects Certification Exam offered by the University Committee on Activities Involving Human Subjects aka UCAIHS (try to say any of that 3x fast!).

It’s all about understanding the ethical way to do research involving human subjects and the certification means you’ve got a good handle on what you need to know. I passed on the 2nd try with an 84 (got a 72 the first time around), so I’m apparently qualified enough to torture people at will. Yay for me!

The Quincunx Via Processing

An assignment for the Crafting With Data class that has been stumping me is how to build the quincunx. I’m at a bit of a disadvantage at this class: I kinda suck at programming and haven’t yet taking PComp, so finding my way around those sort of assignments in a way that won’t have me tearing my hair out usually involves me doing…another survey.

This time around though I’ve decided to kill two birds with one stone a bit and use the quincunx dilemma as my ICM final. At this rate, I can make the pegs, falling balls, and pile at the bottom of the screen (sorta) seemlessly in Processing, but the hard part for me is the collision factor. I’m also hoping that I can actually mimic the random bell curve — which technically should be a given if the construction is correct. Three more weeks until my (hopeful) success!

An example of a quincunx in real life: the Plinko board on The Price is Right (flash example via Drawlogic).

ITP Extracurricular Survey Results

I sent out my survey to The List and got about a 9% response (19 out of just over 200). Not very surprisingly there were a fair amount of ex-high school writers, theater kids, and musicians, but I was surprised to see that 8/17 responders played sports. Perhaps a bit of bias on my part.

It would’ve been interesting to see if certain patterns would’ve emerged with more responses. This sample was too small to be very conclusive. Results in chart forms below.

High School Extracurriculars Survey

The winner of the Discovery Seeker Ideas that I presented to class was a survey exploring high school extracurriculars of ITP students.

I made a web-based survey and sent a test to the class. Problems that I ran into even creating it was finding myself double-guessing every option I chose because I felt overly conscious of making it flexible enough to catch every sort of response. The service has its own limitations of how many questions you can create and responses that can be collected, so the process was frustrating.

The test feedback from the class was really helpful because I was able to fix things that I had totally overlooked while being overly neurotic. Like better ways to phrase questions and options that were too limiting.

I’m hoping to tweak some more over the next couple of days before blasting the survey to the whole program. I’m really looking forward to gathering and analyzing the data.

Coin Toss Experiment

This week’s homework assignment was to do a coin toss 1oo times. Caveat for my group was that I had to pretend to do it. I wrote Heads and Tails in a spreadsheet and alternated arbitrarily.

Results

My results were:
Heads: 55
Tails: 45

Discovery Seeker Ideas

Here are six ideas I’ve been kicking around the potential to do some discovery on:

1) A rehash of the city data experiment, but this time to explore the road I didn’t travel down: population loss and its relation to a city’s education levels.

2) My original idea for the data gathering experiment: when is drunk o’clock?
Some background: I have a lot of bartender friends and I call the time when the drunk to sober ratio is bad (or good, depending on your perspective), the noise levels rise, and the place becomes insanely busy/annoying is drunk o’clock. I want to measure things like crowd numbers, sound levels, and do some social observation to have a bit of a definitive idea on that.

3) Create a survey to administer to NYU students, along the lines of choose your own adventure. I want to see if majors correspond to the responses along the lines of 7 Habits of Highly Innovative People.

4) Similar to 3, but administering a personality test instead.

5) Study on how Facebook, etc. encourage news/memes to move around virally. Partially inspired by the Balloon Boy story of the other week where I was able to learn of and follow the whole story unfolding through friend status updates.

6) Survey of ITP students extracurriculars from high school. I found that learning those gives you interesting insight into people.

Countries In Africa Survey

I surveyed eight people about their guesses on the percentage of countries in Africa in the United Nations.There were 2 groups of 4 with the percentage reference point as 65% or 10%.

The results:

Survey Results

What I found interesting was how curious people were about what the right answer was. (I looked it up and it was about 27-28%.) I also have pretty smart friends because most were within 10%.