BA 762 Research Methods

Bad science?

2009-08-31T13:13:00.000-07:00

Some reactions to the recent Science issue spotlighting some work in network science.

http://www.scientificblogging.com/adaptive_complexity/networks_are_killing_science

Room change for BA 762

2009-08-30T14:19:00.000-07:00

Gentlefolk, we have changed rooms. From now on, class meets in B&E 446. Well, except for when we don't, which is:

Sept4. Meet in B&E 452
Dec 11. Meet in B&E 209

Predicting baseball injuries

2009-07-07T17:20:00.000-07:00

The article doesn't say much, but still interesting:

http://www.nytimes.com/2009/07/08/sports/baseball/08injuries.html

Getting used to technology

2009-07-07T06:55:00.000-07:00

On getting used to technology.

NSF Workshops on Anthropological Methods

2009-02-14T08:23:00.000-08:00

Here are some upcoming workshops along with application deadlines. Two of them are Feb 15. More info on all available on the Methods Mall, at http://www.qualquant.net/training/

NSF's 2009, one-week summer courses on research methods in cultural anthropology (for those with the Ph.D.) Application deadline: Feb 15
NSF's field training program in Bolivia (for graduate students). Application deadline: Feb 15
NSF's three-week summer training program in research design for cultural anthropology (for graduate students). Application deadline: April 1

Labrador decision model

2009-01-18T15:34:00.000-08:00

Math whiz finds fame by calling it for Obama

2008-11-10T19:53:00.000-08:00

Today's New York Times has an article on Nate Silver, the fivethirtyeight.com guy. It talks about how he predicted Obama would win the democratic nomination and 833 Super-Tuesday delegates at a time when Clinton was up by 20 percantage points in the polls. Obama actually won 848 delegates on Super-Tuesday.

Then Silver predicted Obama would beat McCain. That was in March. Here's the full text.

Forecasting the election using simulations

2008-11-03T18:39:00.000-08:00

Here's a cool site that forecasts election results using regression and simulation techniques. Note that many of the percentages you will see on the site are probabilities of winning, rather than percentages of the vote. The site is http://www.fivethirtyeight.com/

McCain's use of text analysis

2008-11-02T11:45:00.000-08:00

The Times is reporting that Republicans offered an Oxford professor $10K to prove that Obama's autobiography was ghost-written by William Ayers, America's favorite Weatherman. Get the full story here.

Obama's success due to marketing science?

2008-10-30T18:41:00.000-07:00

A recent article suggests that Obama's campaign has benefited from the work of a psychologist, Drew Western of Emory University. The full article is here.

Science without theory

2008-10-28T10:06:00.000-07:00

Kevin Kelley's post on doing science without theory is interesting. I disagree with it wholeheartedly, but is worth reading.

Full text is here.

How people perceive the world

2008-10-28T09:46:00.000-07:00

In yesterday's NY times, David Brooks has a nice piece on decision-making. It dovetails well with the Professional Vision paper that you read last week, along with our discussion during Emily's talk. One of the comments he makes is this: "Perceiving a situation seems, at first glimpse, like a remarkably simple operation. You just look and see what’s around. But the operation that seems most simple is actually the most complex, it’s just that most of the action takes place below the level of awareness. Looking at and perceiving the world is an active process of meaning-making that shapes and biases the rest of the decision-making chain."

Read the full text here.

Distinguishing fact from fiction

2008-10-26T08:37:00.000-07:00

The central problem in empirical research is distinguishing fact from fiction (unless you are a postmodernist, in which case the task is to distinguish fashionable, status-enhancing fiction from embarrassing fiction). The Snopes site tries to track and verify urban legends.

It's interesting to search for events and names of current interest, such as "Obama".

Unusual economic indicators

2008-10-19T06:43:00.001-07:00

A long time ago, it was noticed that hemlines fluctuated with the economy: the higher the hemlines, the stronger the economy (I'm not hinting, just saying ...). Now, a number of other correlations have been found, from the sale of laxatives to the tempo of hit songs. See the full article in today's New York Times.

e and pi go out for a date

2008-10-18T08:49:00.000-07:00

Change in direction for the class

2008-10-16T11:32:00.001-07:00

I realized after the last class (on factor analysis) that I was doing the class a disservice. The mathematical underpinnings of measures of association, regression, factor analysis etc are not what you guys need most. There are kind of three ways of teaching a course like this.

Approach 1. In-depth examination of research methods from a formal, mathematical perspective. So formal measurement theory, vectors and matrices, how correlation really works, the underlying math of factor analysis, and so on.

Approach 2. In-depth, hands-on workshop on how to use the main techniques of data collection, construction and analysis in the social sciences. Mostly about the software and the actual nuts and bolts.

Approach 3. Survey of the components of a research project. Exposure to key concepts, such as measurement, validity, reliability etc, as well data collection techniques and data analysis techniques. Emphasis on what they are and what they are used for, but not how they work.

I intended the course to be a combination of Approach 1 and Approach 2. But now I am going to reorganize things to be about 70% Approach 3, and 30% Approach 2. My sense of aesthetic purity says go 100% Approach 3, but I can't resist showing you some tools you can use to analyze real data. It is too much fun.

Of course, it is hard to let go of the math stuff too. So my approach is going to be "if you don't ask, i won't tell". But I hope you do ask about the mathematical nuts and bolts underneath the techniques, and I will explain.

Anthroplogy Methods Mall

2008-10-16T08:15:00.000-07:00

The 2009 Anthropology Methods Mall is online. This site has info about four, NSF-supported opportunities for methods training in cultural anthropology.

1. Now in its fifth year, the SCRM (Short Courses on Research Methods) program is for cultural anthropologists who already have the Ph.D. Three five-day are offered during summer 2009 at the Duke University Marine Lab in Beaufort, North Carolina.

Behavior Measurement (Raymond Hames and Michael Paolisso) July 13–17, 2009
Network Analysis (Jeffrey Johnson and Christopher McCarty) July 20–24, 2009
Systematic Techniques for Gathering and Analyzing Video Data (Elizabeth Cartwright and Jerome Crowder) July 27–31, 2009

2. Now in its 14th year, the SIRD (Summer Institute on Research Design) is an intensive, three-week course for graduate students in cultural anthropology who are preparing their doctoral research proposals. The 2009 course runs from July 13–31 at the Duke University Marine Laboratory. (Instructors: Jeffrey Johnson, Susan Weller, and H. Russell Bernard)

3. Now in its sixth year, the SFTM (Summer Field Training in Methods) program in Bolivia is open to graduate students in cultural anthropology. This course involves five weeks of fieldwork in the Bolivian Amazon from June 8-July 13. (Instructors: Ricardo Godoy, William Leonard, Victoria Reyes-Garcia, Thomas McDade, Clarence Gravlee, J. Richard Stepp, and Susan Tanner).

4. The WRMA (Workshops in Research Methods in Anthropology) program offers one-day workshops in conjunction with national meetings of anthropologists. Two workshop will be offered at the meetings of the American Anthropological Association in November.

Introduction to social network analysis. (Instructors: Jeffrey C. Johnson and Christopher McCarty). November 20, 2009.
Demystifying SPSS®: Anthropological Data Management and Analysis Made Easy . (Instructors: William Dressler and Kathryn Oths). November 22, 2008.

Full details on all these opportunities at the Methods Mall, http://www.qualquant.net/training/.

H. Russell Bernard
Professor of Anthropology, Emeritus
University of Florida
Editor, FIELD METHODS
http://www.qualquant.net/FM/

What's the difference between sense and nonsense?

2008-10-14T09:42:00.000-07:00

Listen to this youtube video.

Artistic Text Analysis

2008-10-13T15:15:00.000-07:00

A beautiful way of visualizing the words in a text. Word size is proportional to frequency. The text in this case was an interview by conducted by Ginny Kidwell and Travis Grosser for a project about the MBA job search process.

For more information, visit the course website or the site that generated the picture, wordle.net

The SMOOT unit of measurement

2008-10-12T12:44:00.000-07:00

The Mass Ave bridge linking Boston to Cambridge is 364.4 Smoots long. Read all about it here.

Geneology of Factor Analysis

2008-10-09T14:13:00.001-07:00

Here's a little bit of history -- graphically presented -- on the development of factor analysis.

Family tree of the developers of factor analysis
Time line

When dumb is smart

2008-10-08T17:58:00.000-07:00

Maureen Dowd recently wrote: "We could, following [Sarah Palin's] strenuously folksy debate performance, wonder when elite became a bad thing in America." [full article]

It's an odd thing, but being smart is bad. It's always been true that in school, it was uncool to be smart (or at least to be into school). But now, being tagged with intellectual is practically like calling someone a communist. Even odder, the proponents of this are both the right and the left. On the right we have people who put creationism on a par with evolutionary theory. The left, we have people who argue that reality is unknowable and "expose" science as a master narrative designed to keep old white men in power.

Physics rap

2008-10-05T08:24:00.001-07:00

Don't like rap? Too much violence and sex? You haven't seen anything yet -- watch this rap on the physics of the new CERN particle collider.

Now that is sex and violence!

Medieval help desk

2008-10-05T08:23:00.000-07:00

Need help learning Vista or Office 2007? Can't figure out UCINET? It has always been this way with information technology -- all the way back to the introduction of the Book. Check out the medieval help desk on YouTube.

Welcome!

2008-10-05T08:14:00.000-07:00

This is a blog for discussing research methods and issues of science and epistemology. It is meant as a companion to BA 762, a PhD level research methods course at the University of Kentucky.