5 of 6 people found the following review helpful
Troy Lynch ([...]),
This review is from: Hands-On Intermediate Econometrics Using R (Hardcover)
I have been using this text primarily as a way to learn R and econometrics. I cannot comment on the level of econometrics, though he seems to cover the core areas. Where it is a valuable resource is in the practitioner space between econometric theory and the immediate application of that using a live application - R.
It is the sort of text that, if you follow step-by-step, should bring you up to speed with R very quickly. Alternatively, it is a great resource for R code that has been tried and tested. As with any new piece of software, you need to bumble around until you get the hang of the core structure of the language: grab code from here and there, use it, apply it to real data and see what happens.
It is more than just an R code resource, a test on intermediate econometrics, and a tool by which to learn R. I think of it as a back door to econometrics via R. A great way to learn econometrics is to actually do it, rather than cogitating over technicalities, theories and rules. Humans learn better by doing (and learning the formal rules along the way) - hence his title: "Hands-on..." Vinod has done a great job for the novice R user and the novice econometrics user.
This text would sit neatly beside "Forecasting: Methods and Applications" (3rd edition) Makridakis, Wheelwright and Hyndman, 1998 and "A Guide to Econometrics" (6th edition) by Peter Kennedy ISBN 9781405182577. The former is a practitioner's introduction to forecasting -- and a lot of fun -- and the latter, by Kennedy, is a great explanatory tool for those wishing to actually understand econometric theory.
Also check out Hyndman's website: [...]. You can obtain all the data from the Makridakis, Wheelwright and Hyndman, 1998 text there, and use the tools in Vinod to run the examples listed in Makridakis, et al 1998. There are also some helpful snippets and worked examples on Vinod's book site: [...]
Hyndman is an editor of the International Journal of Forecasting: [...] He and others (e.g. Makridakis, who is now running the M4 competition) have for many years been great writers on the methodology of forecasting. Armstrong, for one, has also done a great service with his exhaustive analysis of the accuracy of forecasts: "Long-Range Forecasting: From Crystal Ball to Computer" (1985), now available for free, on-line, at [...]). Okay, these guys are statisticians, but the beauty of all of this is their tenacity in blowing the whistle on many so-called tried and proven techniques employed for forecasting.
The conclusion in Makridakis, Wheelwright and Hyndman (1998) is that simple and combination models are more accurate than complex ones. This blows the lid on complex econometric techniques. Armstrong also notes a similar conclusion in his 1985. (See also his Principles of Forecasting Handbook: [...] The caveat is that simple and combination is not the solution for all applications.
However, as an Austrian economist, I find it amusing that econometricians are empiricists in search of a theory: the next tweak will resolve the issue. That said, the aforementioned forecasters are arguing than simple statistics (without the use of multivariate explanatory variables) seem to perform better at minimizing error terms! It seems that standard marginal economic analysis is shunted aside by micro- or macro-econometrics; but the latter is, in turn, swiftly brought to task by the statistical forecasting community.
Buy Vinod and Kennedy to learn R and econometrics; and treat yourself to Makridakis, Wheelwright and Hyndman (1998) and Armstrong (1985) to learn forecasting. I read recently, I think by Armstrong, that Makridakis, Wheelwright and Hyndman (1998) is still considered the bible for forecasting - not econometrics, but forecasting - though Hyndman, on his website, notes that they are not producing a new edition; he is attempting to put together an on-line edition at[...].