SI 544 Introduction to Statistics and Data Analysis


SI 544 home

Readings, assignments, etc. will be posted to the course ctools website

problem sets

software tools for the class

Lada Adamic


Winter 2009:

Lectures will be
Tuesdays and Thursdays
from 9:00 to 10:30 am.
Location WH 409

Discussion section Thursdays 6:30-7:30 WH409

Office hours:
Lada: Mon 5-6pm WH 3082
Tracy: Thurs 7:30-9:30 WH409

544 student mini project showcase



Students are tasked with completing group projects of small scope (producing a 3 page report). The topics are entirely chosen by them, the main task is to carefully gather data and do an appropriate statistical analysis.

Click here for the syllabus of the course.

Mac users more likely than PC users to stay at home on Friday nights even though supposedly "on the prowl"

A study conducted by School of Information masters students Jon Cohen, Elaine Engstrom, Noah Liberman and Laura Rodian carefully coded stereotypes promoted by Apple TV commercials. Then they surveyed SI masters students to see which, if any of these stereotypes hold up. Very few did: PC users ran virus checkers more often, and Mac users used design software more, but little else, besides this: Mac users were more likely to stay in on Friday nights, even though they were less content to be single. Read more...

Earn more, spend more time with family.

Think making the big bucks means losing work/life balance? Think again. According to a study by Elizabeth Bedford, Chun-Yuen Tend, and Rebecca Yu, those who earn more actually spend slightly more time on average with their families, and this is more true for men than women. By carefully analyzing the 2007 American Time Use Survey data, they also found that those with more children spend more time on family, but the difference is more pronounced for women. Those with no jobs at all spent more time with family, but for women this was true of both those seeking work and those not in the labor force, while for men it only held for those seeking work.

Countries with higher proportion of tourism income less polluted, study finds.

Greenpeace may have raised an alarm about tourists polluting the countries they visit, but this study shows that the places where the tourism industry is big are still less polluted than other countries. After carefully gathering country level data on proportion of GDP from tourism, and several air pollutant concentrations, Jackie Cerretani, Pratibha Bhaskaran, Yesook Im, and Silhee Sung found a moderate negative correlation between pollution and tourism income.

NBA all star selection not just a popularity contest

Sometimes an basketball team having an all-star player still hits a losing streak. Does having all stars on a team increase the winning percentage, the odds of making it to the playoffs, attendance? Yes, yes, and yes, find study authors Ke Sun, Chris Wieland, Saurabh Koparkar and Kevin McHenry. Though initially skeptical about whether the contest was about little more than popularity, they find that points scored by a player are strongly predictive of whether that player will become an all star.

Weight and height matters in tennis players' performance

Zhenan Hong, Chih-Ming Yi, and Wentao Wang analyzed data on top male and female tennis players. Their findings? Men who weight close to 85kg tend to score more tournament points, while for women, the taller players outperform shorter ones.

Video game sales recession-proof? Not so, study finds

It's been said that when the economy goes south, hollywood does just fine - people like to escape by watching movies. Wenbo Wang, Daniel Fehrenbach, and Tao Dong wanted to see if the same was true of another entertainment industry - the videogame industry. They find that changes in videogame sales appear to lag the personal income by a year. Although game sales do not seem immune, the past few years, which have seen the dramatic rise in popularity of gaming consoles, may yet ride out this recession just fine.

Where are all the (stupid, intoxicated and cold) men?

Annie Fang, Eric Garcia and Adrienne Klum hung out late at night in Yahoo romance chatrooms. While there, they gathered some state-level participation stats (what did you think they were doing there ;)?), and matched it against other state level data: av. temp., beer consumption, wine consumption, education level. Their findings? Cooler states with higher beer consumption but lower education levels seek romance on Yahoo!

Students don't need computer labs because they own their own computers, right? - not so...

Having gotten a hold of a survey conducted by ITCS, study authors Rodger Devine, Sebastian Encina, John Levandowski, and Jamie Niehof examine whether computer labs are dinosaurs, or very much needed resources. In particular they find a slight positive association between computer ownership (laptop or desktop) and frequency of use of computer labs. They also find that younger students are more likely to use the labs.

Popularity of movies reflected in quotes

You may think that critic ratings would be a good predictor of movie success at the box office. Not so, find study authors Matt Adamo, Neha Kumari, Tanuj Shah, and Matt Simmons. Based on gather movie ratings and box office grosses off of the Rotten Tomatoes site, and pairing that with quotes retrieved from the internet movie database (imdb), they find that a far better indicator of box office grosses are the number of quotes users submit on the imdb website.

Bond movie fails at the box office. Time to change actors?

Graham Dethmers, Aalap Doshi, Jason Stewart and Caroline Yee took advantage of data on the long running 007 franchise to see whether the box office responds differently to different actors, and it does. There is also an apparent jump in box office revenues when a new Bond actor is hired - though hopefully the franchise will continue for a few more decades so these jumps become statistically significant :).

Sports activities depend on age and income level

Chen Huang, Yang Liu, and Ning Wang use the American Time Use Survey data to study how sport activities depend on age and income level. There are some activities, such as walking - favored by the older folk. Income makes a difference too - basketball is more popular among lower income and younger people, while fitness training is popular among middle and high income groups.

North or South - which schools party harder?

Ben Malley, Jeremy Canfield, Josh Steverman, and Megan McGlynn put the southern party school stereotype to the test. They use self-reported data on alcohol and drug-related on-campus arrests from all postsecondary schools participating in federal student aid programs (who are required to report these stats). They find that it is actually the northern schools with higher rates of alcohol and drug-related arrests, but also question the inherent bias in the data (some schools may combat these offenses more aggressively than others).