Innovations in Editing

I've posted in the past (here and here) about some of my experiences as an academic journal Editor. It has its ups and downs, for sure, but ultimately it's a rewarding job.

Preston McAfee (Yahoo! & Caltech) is currently the Editor of Economic Inquiry, where he's introduced some important innovations into the editorial process. He was formerly Co-Editor of American Economic Review. As well as being a highly respected economist, Preston has a wonderful way with words.

His thoughts on journal editing make excellent reading, for seasoned academics and newcomers alike. I particularly recommend Preston's piece in the American Economist last year - here.

Reference

McAfee, R. P. (2010). Edifying editing. American Economist, 55, 1-8.

© 2011, David E. Giles

Visualizing Random p-Values

Here's a follow-up to yesterday's post on Wolfram's CDF file format. In an earlier post (here) I discussed the fact the p-values are random variables, with their own sampling distribution.

A great way of visualizing a number of the points that I made in that post is to use the CDF file for the Mathematica app. written by Ian McLeod (University of Western Ontario). You can run it/download it from here, using Wolfram's free CDF Player (here).

You'll recall that a p-value is uniformly distributed on [0 , 1] if the null hypothesis being tested is true. Using Ian's app., here is an example of what you see for the case where you're testing the null of a zero mean in a normal population, against a 2-sided alternative:

The null hypothesis is TRUE. The sample size is n = 50, and the simulation experiment involves 10,000 replications. You can see that the empirical distribution is approaching the Uniform true sampling distribution.

When the true mean of the distribution is 2.615 (so the null hypothesis is FALSE), the sampling distribution of the (two-sided) p-value looks like this:

If you're not sure why the pictures look like this, you might want to take a look at my earlier post, ("May I Show You My Collection of p-Values?")

© 2011, David E. Giles

Interactive Statistics - Wolfram's CDF format

Many of you will be familiar with Wolfram Research, the company that delivers Mathematica, among other things. Last month, they launched their new Computable Document Format (CDF) - it's something I'm going to be using a lot in my undergraduate Economic Statistics course.

Here are a few words taken from their press release of July 21:

Themes in Econometrics

There are several things that I recall about the first course in econometrics that I took. I'd already completed a degree in pure math. and mathematical statistics, and along with a number of other students I did a one-year transition program before embarking on a Masters degree in economics. The transition program comprised all of the final-year undergrad. courses offered in economics.

As you'd guess, the learning curve was pretty steep in macro. and micro, but I had a comparative advantage when it came to linear programming and econometrics. So things balanced out - somewhat!

That Darned Debt!

What with the recent S&P downgrading of the U.S., and the turmoil in the markets, it's difficult to watch the T.V. or read a newspaper without being bombarded with the dreaded "D-word". I was even moved to pitch in myself recently by posting a piece about the issue of units of measurement when comparing debt (a stock) with GDP (a flow).

Yesterday, Lisa Evans had nice post titled 20 Outlandish and Informative Ways to Illustrate the U.S. National Debt. I think you'd enjoy it!

Lisa runs a site called Masters in Economics. It's devoted to assisting students who are thinking of undertaking a Masters degree in Economics as a stepping stone into a career in Economics. Programs at this level are more widely available than you might have thought.

I'm hoping that Lisa will be able to expand her database to include the many programs at this level in Canadian schools.

Meantime, keep an eye on her site and watch for her future posts.

© 2011, David E. Giles

Flip a Coin - FIML or 3SLS ?

When it comes to choosing an estimator or a test in our econometric modelling, sometimes there are pros and cons that have to weighted against each other. Occasionally we're left with the impression that the final decision may as well be based on computational convenience, or even the flip of a coin.

In fact, there's usually some sound basis for selecting one potential estimator or test over an alternative one. Let's take the case where we're estimating a structural simultaneous equations model (SEM). In this case there's a wide range of consistent estimators available to us.

There are the various "single equation" estimators, such as 2SLS or Limited Information Maximum Likelihood (LIML). These have the disadvantage of being asymptotically inefficient, in general, relative the "full system" estimators. However, they have the advantage of usually being more robust to model mis-specification. Mis-specifying one equation in the model may result in inconsistent estimation of that equation's coefficients, but this generally won't affect the estimation of the other equations.

The two commonly used "full system" estimators are 3SLS and Full Information Maximum Likelihood (FIML). Under standard conditions, these two estimators are asymptotically equivalent when it comes to estimating the structural form of an SEM with normal errors. More specifically, they each have the same asymptotic distribution, so they are both asymptotically efficient.

Being Normal is Optional!

One of the cardinal rules of teaching is that you should never provide information that you know you're going to have to renege on in a later course. When you're teaching econometrics, I know that you can't possibly cover all of the details and nuances associated with key results when you present them at an introductory level. One of the tricks, though, is to try and present results in a way that doesn't leave the student with something that subsequently has to be "unlearned", because it's actually wrong.

If you're scratching your head, and wondering who on earth would be so silly as to teach something that has to be "unlearned", let me give you a really good example. You'll have encountered it a dozen times or more, I'm sure. You just have to pick up almost any econometrics textbook, at any level, and you'll come away with a big dose of mis-information regarding one of the standard assumptions that we make about the error term in a regression model. If this comes as news to you, then I'll have made my point!

The Article of the Future

The standard format for academic journal articles is pretty much "tried and true": Abstract; Introduction; Methodology; Data; Results; Conclusions. There are variations on this. of course, depending on the discipline in question. When it comes to journals that publish research in econometrics, it's difficult to think of innovations that have taken advantage of developments in technology in the past few years.

O.K., so you can follow your favourite journal on Twitter or Facebook - but when you get to the articles themselves, do they look that much different from, say, ten years ago? Not really.

You'd think that econometrics journals could be a bit more exciting. The content is always a blast, of course, but what about the way it's presented? Apart from moving from paper to pdf files, we haven't really come that far. Yes, in many cases you get hyperlinks to the articles listed in the References section, which is fine and dandy. But is that enough?

When we undertake the research that leads to the articles we use all sorts of data analysis and graphical tools, whichever econometric or statistical software we're wedded to. Yet, these powerful tools are left pretty much on the sideline when it comes to disseminating the information through the traditional peer-reviewed outlets.

Are there any glimmers of hope on the horizon?

In a recent electronic issue of their Editors' Update newsletter, the publishing company, Elsevier, discussed their so-called "Article of the Future" Project. It's worth looking at. In particular, it could get you thinking about how we can raise the bar a little when it comes to publishing new research results in econometrics. For example, see the interactive graphics in the "Fun With F1" article.

A lot of us have some pretty strong views about the pricing policies of academic journals, and about the extent to which the flow of scientific information should be commercialized. I have no affiliation with this particular publisher, but it's refreshing to see what they're up to in this regard. Hopefully there's more of this going on elsewhere.

© 2011, David E. Giles

Galton Centenary

The words "regression" and "correlation" trip off our tongues on a daily basis - if not more frequently. Both of them can be attributed to the British polymath, Sir Francis Galton (1822 - 1911). I've blogged a little bit about Galton rpreviously, in The Origin of Our Species.

To commemorate the centenary of his death on 17 January 1911, statisticians are honouring Galton's impressive contributions this year. Putting aside Galton's promotion of eugenics, there is still much to celebrate. Perhaps the most comprehensive source of information about his work and influence is at http://galton.org/. This site includes, among other things, copies of all of his published work - much of which is difficult to obtain elsewhere these days.

Not surprisingly, the Royal Statistical Society has been paying special to Galton this year. Among other things there have been some interesting items in their Significance magazine. I'd especially recommend the pieces by Graham Wheeler, Tom Fanshawe and Julian Champkin. In the last of these, look for the link to a BBC radio talk on Galton, by Steve Jones of the Galton Laboratory at University College London!

Finally, if you're looking for inspiration - and who isn't(!) - Galton's own account of his discovery of correlation and regression (originally termed "reversion") makes interesting reading. Titled "Kinship and Regression", you can find it here.

© 2011, David E. Giles

Moving Average Errors

"God made X (the data), man made all the rest (especially ε, the error term)."

Emanuel Parzen

A while back I was asked if I could provide some examples of situations where the errors of a regression model would be expected to follow a moving average process. 

Introductory courses in econometrics always discuss the situation where the errors in a model are correlated, implying that the associated covariance matrix is non-scalar. Specifically, at least some of the off-diagonal elements of this matrix are non-zero. Examples that are usually mentioned include: (a) the errors follow a stationary first-order autoregressive (i.e., AR(1)) process; and (b) the errors follow a first-order moving average (i.e., MA(1)) process. Typically, the discussion then deals with tests for independence against a specific alternative process; and estimators that take account of the non-scalar covariance matrix - e.g., the GLS (Aitken) estimator.

It's often easier to motivate AR errors than to think of reasons why MA errors may arise in a regression model in practice. For example, if we're using economic time-series data and if the error term reflects omitted effects, then the latter are likely to be trended and/or cyclical. In each case, this gives rise to an autoregressive process. The omission of a seasonal variable will general imply errors that follow an AR(4) process; and so on.

However, let's think of some situations where the MA regression errors might be expected to arise.