Those looking for either introductory or mid-level instructional on-line resources should visit UCLA's idre site. The site's resources are both deep and wide. Particularly helpful is an array of on-line classes and workshops (some with accompanying videos) that span the leading statistical software packages and include numerous, discrete topical areas. Well worth a look and kudos to UCLA for making it publicly available.
Though most of our stats software package-related posts skew towards Stata, we remain mindful that folks' tastes vary across the most popular packages (e.g., Stata, SPSS, R, SAS). With this in mind, we recently stumbled across a helpful and quite user-friendly SPSS resource. Mounted by SPSS' new owner (IBM), the "Case Studies" tab provides hands-on examples of how to perform various types of
statistical analyses and interpret the results. The site walks users through (using quite helpful "screen-shot" slides) a suite of statistical tests found in SPSS' various stats packages. (For one example, factor analysis is explained here.) Worth a look for those using SPSS.
While noting that Stata does not shy away from holding its own initiatives up to data should not surprise, a summary report on the initial six months of Stata's YouTube Channel usage might interest readers. If there is any surprise in a summary (and descriptive) blog post (here), perhaps it's the degree to which this project is succeeding. What Stata is doing is certainly generating interest (thus far, anyway) and I assume we'll see similar efforts by others in the near future.
Information on two separate (Main and Advanced) Causal Inference Workshops at Northwestern Law School this summer follows.Both workshops will be taught by world-class
causal inference researchers. Registration for each is limited to 100 participants.
Main workshop: Monday – Friday,
June 24-28, 2013
Advanced workshop: Monday - Wednesday,
August 12-14, 2013.
For information and to register for either or both workshops: (click here)
Overview and Target Audience: Most empirical methods
courses survey a variety of methods. We will begin instead with the goal of
causal inference, and discuss how to design research to come closer to that
goal. The methods are often adapted to a particular study. Some of the methods
we will discuss are covered in PhD programs, but rarely in depth, and rarely
with a focus on causal inference and on which methods to prefer for messy,
real-world datasets with limited sample sizes. Each day will include with a
Stata “workshop” to illustrate selected methods with real data and Stata
code. We will
assume knowledge, at the level of an upper-level college econometrics or similar
course,of multivariate regression, including OLS, logit, and probit;
basic probability and statistics including conditional and compound
probabilities, confidence intervals, t-statistics, and standard errors; and some
understanding of instrumental variables.
Advanced Workshop Overview and Target
advanced workshop seeks to provide an in-depth discussion of selected topics at
the causal inference research frontier.Our target
audience is empirical researchers who are familiar with the basics of causal
inference (from our main workshop or otherwise), and want to extend their
While I've discussed the Stata instructional videos previously, recent notable additions to the video collection include a helpful 3-part series on Introduction to Margins. The entire collection of instructional videos can be found here.
From the Stata blog: "StataCorp now provides free tutorial videos on StataCorp’s YouTube channel." A direct link to Stata's YouTube "channel" (here). This is, of course, a wonderful move by Stata. While the current collection of videos is limited, more are "forthcoming." My only quibble is that the videos are broken into quite small (and short) discrete sessions (thus far, 24 separate videos accounting for just under 2 hours of instruction). Any quibbles aside, well worth a look.
Yale Law School will begin offering a Ph.D in law, with the first incoming class arriving in Fall 2013. I learnd of this from a WSJ story; the Yale press release is here. From the latter:
"Because the level of the scholarship expected of entry-level law professors has risen quite dramatically, increasing numbers of law professors now pursue Ph.D.’s in allied disciplines like economics, history, philosophy, or political science. Because such disciplines train students in standards and questions that are different from those of the law, the natural next step for the legal academy is to create our own Ph.D. program that can focus on the questions and practices of the law itself. Students obtaining a Ph.D. in law may, of course, engage in interdisciplinary studies, but their work will be anchored in the framework of legal scholarship." (Dean Robert Post)
The program is clearly designed to prepare individuals for careers on law faculty; and, interestingly, the program is only open to individuals who have already received a J.D. This makes it an interesting move by Yale, both from a "possible pool of candidates" perspective and (more broadly) given some of the other changes in law schools recently (and their likely knock-on effects for faculty hiring). I'll have to mull this last bit over a bit more before commenting further, but my knee-jerk reaction is that this will further polarize the ranks of law school faculties with respect to credentals, focus, etc.
In related news, Gordon Silverstein, a political scientist formerly at (inter alia) Berkeley, will assume the position of Assistant Dean of Graduate Programs at YLS, effective Monday. He'll be in charge of the new Ph.D., as well as existing the LL.M. and J.S.D. programs. Congratulations, Gordon!
For any students (or faculty, for that matter) contemplating a summer research project, Andrew Gelman (Columbia--Poli Sci) offers quite helpful, quick advice on how to approach writing up a research paper (here).
Always-provocative lawblogger Vivia Chen's recent post "Too good for Big Law" has drawn a fair bit of attention. She notes work by our very own Bill Henderson (using data from the NLJ) suggesting that associate to partner odds -- that is, the ratio of Biglaw associates hired from a particular law school to Biglaw partners from the same school -- for many elite / T14 schools are lower than those for lower-tier places.
Bill responded with his usual meticulousness at the Legal Whiteboard, offering the "statistician edition" of the Careerist post. After a suitable discussion of the potential caveats, he does a nice job of rounding up a range of possible explanations for this finding. At Above the Law, David Lat casts his vote for "Selection Effects," "Inter-Generational Privilege," and "A Better Plan B" as the most likely drivers among among BH's explanations. And Bruce MacEwen at Adam Smith Esq. lays out the data in all its glory, and has some thoughtful ruminations on why we might not want to take the results too seriously, and why (and how) we might. There's also an exposition at TaxProfBlog. For its part, the NLJ offers an interactive chart that puts some of their data in visual form (including tuition for the schools in question).
Of course, I just couldn't help myself from digging into the data a little. All of the commenters above make the point that, for this comparison to work, the numbers cannot be changing much from year to year; for example, at one point Henderson says "I analyzed this same data four years ago and got essentially the same results." So I found the same NLJ data from 2010, and did a little comparison.
The vertical axis correspond to the (2011) percentages listed at Bruce M's post; the horizontal one is the same numbers for 2010. (My UH number is a bit different from his, but the others are spot-on). Red "Xs" are the "Top 14" law schools, while black dots are the rest; I labeled a few of the outliers. The data are noisy: the simple (Pearson) correlation between the two years is 0.32. But the correlation is driven almost entirely by the difference between elite and non-elite schools: the correlation among non-elite schools only is 0.06, while that among elite schools only is a paltry 0.02. Moreover, the aggregate differences between Top 14 and non-Top 14 schools are large, averaging 17.2% vs. 28% in 2010 and 15.9% vs. 40.5% in 2011.
So, while the data are noisy, the larger pattern still holds: More elite schools have consistently lower "partner yield rates" than do less elite, tier-one schools. While this doesn't begin to get at the various reasons why this might be happening, it does lend some support to the idea that there's something here worth looking into.
Over at Concurring Opinions, Dave Hoffman's (Temple) fascinating post features a simple graphic that speaks, quite literally, for itself and potentially implies a lot for how first year civil procedure is traditionally structured and presented. Well worth a look.