Lecture, three hours. Preparation: familiarity with data analysis software (e.g., R, Excel, MATLAB, Stata) and/or programming experience. Enforced requisites: courses 101, 103, 103L. Survey of theory and application of time-series methods to forecasting in economics, business, and government. Topics include modeling and forecasting trend, seasonality, and cycles. Discussion of stochastic trends, volatility measure, and evaluation of forecasting techniques. Hands-on approach to real-world data analysis methods widely used by economists and other professionals. P/NP or letter grading.
3.3 / 10
6.7 / 10
0.0 / 10
I genuinely don't understand why Econ students think Rojas is a good professor, because this class was dogshit. Rojas isn't good at explaining concepts, his lectures are unengaging, and the lectures do hardly anything to prepare you for the homework and group projects. The single shining upside for this class is that it is, hands-down, the most practical Econ class you will ever take if you want to get a job. Learning how to forecast in R is very useful, I just wish Rojas wasn't so shit at teaching it.
The homework/group projects are assigned roughly weekly (except for the first week and the midterm week) and require several hours to complete. They're all focused around coding in R, and you have to figure out a lot of it yourself. Rojas tries to demonstrate some R coding in lecture, but it's completely unhelpful. This is the only class where I've ever actually wished for a discussion section, and it doesn't have one. The feedback on homework/projects is generally unhelpful, and gives you little chance to improve.
Rojas assigns three (yes, three!) textbooks for the class, with hours of optional weekly reading for each of them. I don't think it's possible to read everything he suggests and still take other classes, so I ignored textbook B and focused mainly on C (Forecasting Principles and Practices 3rd edition, by Hyndman and Athanasopoulos) and some of textbook A. If you only read one of the textbooks, I recommend FPP 3rd edition, as it's the clearest and the best resource for doing the homework. I wouldn't recommend skipping the reading entirely, as Rojas can't lecture for shit.
The midterm was online, and involved doing coding analysis questions live, which was a time crunch. The final was in-person, and was the worst final I've had in any Econ class (worse than anything in the 11/101/102 series). Rojas asked questions on every single concept ever mentioned in lecture, no matter how briefly (like models that never appeared on the slides, and which he never demonstrated in R, but only spoke about for 20 minutes). He also asked inane questions about the specifics of model implementations in R, which shouldn't really be relevant to the final (you can look those up easily in the R documentation if you ever forgot). At least we got 4 pages of "cheat sheets" for the final, but God help you if you didn't take excruciatingly detailed notes in lecture.
Rojas seems to focus the lectures as if you took 104 with him the quarter right before. He constantly referenced his 104 slides, and the class as a whole felt more like an extension of 104 (which I hadn't taken). 104 isn't listed as a prerequisite for this class, but maybe it should be. Then it would be less of a shitshow.
I took Econ 147 at the same time as this class, and that was a good pairing. The second half of 147 covers introductory time series concepts, and some of the same models (like ARMA, ARCH/GARCH) and concepts (covariance stationarity, white noise processes) covered in 144. It also has a very small amount of coding in R. If you can take 104 and 147 before 144, I highly recommend it, even though they're not prerequisites. The class is doable without them (I still got an A), but it could probably be made much easier if you prepare for it. The Econ department ought to update the prerequisites so students don't make the same mistake I did.
Showing 1 review
Has a group project
Attendance not required
Finals week final
100% recommend the textbook
Textbook information not available.