Curated Collection of Teaching Materials
Applied Linear Models (STAT3012 at USYD)
So I'm trying something new for teaching... Totally redesigned the slides, added a "Fun fact" to pay a homage to Ronald Fisher, and getting down with foundational concepts of experimental design (as per Bailey 2008). Maybe I should hold a short trivia at the end? 🤔 #statsed pic.twitter.com/bxlFUq8Db0
— Emi Tanaka 🌾 (@statsgen) May 10, 2018
How many people know that R was developed in New Zealand @statsauckuni by Ross Ihaka and Robert Gentleman? Tomorrow I'm going to teach factors in R. Factors can be great but it's been the source of my error in my code many times 😅 #statsed #statistics pic.twitter.com/Oq8l7OkZld
— Emi Tanaka 🌾 (@statsgen) May 13, 2018
Do you have **replication** or **repetition**? It's often that pseudo-replication gets mixed up as real replication. I'm going to tackle teaching all things replication tomorrow with nested designs where so many things can go wrong in the analysis... #statsed #statistics pic.twitter.com/Att3INBUmh
— Emi Tanaka 🌾 (@statsgen) May 15, 2018
One analysis of incomplete block design is the intra-block analysis where treatment effects are adjusted for block effects. There is some information in inter-block though! Recovery of inter-block information will be later when we start mixed models 😉 #statsed #statistics pic.twitter.com/rFa4SKwVBC
— Emi Tanaka 🌾 (@statsgen) May 19, 2018
The presence of concomitant variables can lead you to a different conclusion if you don't take them into account. Tomorrow I teach ANCOVA - the combination of regression models with ANOVA models. #statsed #statistics pic.twitter.com/8ZZ6s7ypFY
— Emi Tanaka 🌾 (@statsgen) May 22, 2018
Phew! I have never been busier 😖 And tomorrow I teach random effects model. What should be a fixed effect and what should be a random effect?... I don't know... try both and see which one gives you the result you want? Random rant over. #statsed #statistics pic.twitter.com/X0TTklrb9A
— Emi Tanaka 🌾 (@statsgen) May 24, 2018
I'm teaching linear mixed models tomorrow and my slides have gotten dangerously dense! There is so much that you can talk about with LMM... starting with the variation of the name... MLM? mixed-effects model. Personally, I prefer LMM. #statsed #statistics pic.twitter.com/Gr9sJNh1sS
— Emi Tanaka 🌾 (@statsgen) May 27, 2018
That infamous quote…
The memefulness inspired by @sarah_romanes ✨ pic.twitter.com/eyclrRkj0b
— Emi Tanaka 🌾 (@statsgen) May 27, 2018
Tomorrow teach about variance component estimation of mixed models. My slides are once again dense! I added some color 🎨 so it's not a dull block of text. The content is exciting though as we get to the mathematical component with REML vs ML estimation! #statsed #statistics pic.twitter.com/YomSLWoTmv
— Emi Tanaka 🌾 (@statsgen) May 29, 2018
Just a reminder that women make contributions too!
Today was the last of my "Fun fact" series as I move to application based lectures. All the historical figures in my "Fun facts" have been men but I don't want to send the wrong message! So I had a plug for @RLadiesSydney #rladies and also @StatSocAus today #statsed #statistics pic.twitter.com/GMGVQvUyyK
— Emi Tanaka 🌾 (@statsgen) May 30, 2018
Just completed my lecture slides in a nick of time 😓 Coincidentally, talking about longitudinal data tomorrow. I'm looking forward to a bit of break soon... #statsed #statistics pic.twitter.com/KdbU9ERIy9
— Emi Tanaka 🌾 (@statsgen) May 31, 2018
Modern-day #statistics is largely developed from individuals in the 1900s working with agricultural data. You can perhaps say the root of statistics is in agriculture, don't you agree @antonwasson ? 😁 Tomorrow I revisit split-plot design and small intro to breeding #statsed pic.twitter.com/7yxBucI2xe
— Emi Tanaka 🌾 (@statsgen) June 3, 2018
Another practical use of the linear mixed model is in small area estimation. Tomorrow, I'm teaching area level (aka Fay-Herriot model) and unit level models. Thanks to @roryt_m for putting me in touch with someone from ABS so I can ask how its used there! #statsed #statistics pic.twitter.com/m7X8CwzhoR
— Emi Tanaka 🌾 (@statsgen) June 5, 2018