This week we had guest lectures by Eric on his research on Yosemite toads and Phillip on statistical tests and earthquake prediction.
Both lectures I found quite interesting and well illustrated the statistical methods used (or misused by the earthquake quacks). In particular I found Eric's description of the process and line of reasoning as a scientist he used to expand on his hypotheses and explore the problem space using different methods very elucidating, as well as the practical concerns he had as a researcher. For example, the way he was able to use the model to identify which sites to gather new data at and then use the new data to validate the model.
Looking at how he used the methods we had learned about abstractly in previous classes to really drill down and attack from different angles his lines of inquiry in two very concrete examples helped me to visualize the frameworks and statistical machinery that we use and how it all fits together.
I enjoyed Phillip's lecture also, although some of the material on tests was review for me it was a good refresher and for the tests that were new to me (F-test and ilk) the material was succinct and informative. I do wish the lecture notes would be posted on BSpace though so we could go over them after class.
Class participation during Thursday's lecture however seemed to fall very short compared to Tuesday for some reason or another, whether it be students' haziness on long forgotten fundamentals from 134/135 or the distractions provided by our casual use of laptops and electronic devices. It seems to me that our class culture is in need of marshalling, both to increase boldness in participation (I'm pretty sure many of us knew the answers to the questions asked) and to undivide our focus from the lecturer, particularly in the kind of scenario where one might only need a laptop to take notes or reference materials related to the content being presented.
Also I would like to make a plug for Piazza as I believe it would help solve many of the problems we are having as a class (getting everyone on the same page and getting everyone to help each other to do so) without some of the limitations of the current communications tools we are using (IRC, BSpace, Github) which are not so well designed for that particular purpose. The way I see it we do not want to cultivate a culture of gathering around the mother bird and waiting for the vomit, which bottlenecks class time in a unproductive way.
I think having a clear grading rubric (percentage-wise breakdowns) would help in properly incentivizing out-of-class preparations, as without knowing the relative importance of certain tasks it becomes difficult to prioritize or find motivation in some cases.