Analysis of Messy Data Volume 1: Designed Experiments, Second Edition

Like what I mentioned in the title, this is a somewhat old textbook, but still a to go source I considered very important!I took the class from the second author Dr. Johnson in 2016, he only went back for one semester after his 2009 retirement because the instructor teaching this course was on sebatical, so I was indeed very lucky. The method he taught the course is pretty much similar to reading the book, in this case, the manuscript is the same as the transcript. So that's why this book is pretty long. Except the place with formulas or derivation, he won't stop. And the text sounds so fluent from his reading, yeah, maybe it's because this is his words.Now I started work in an ag company, and yes this becomes more important not only in the design phase, but also in the details of the example. Dr. Johnson and Milliken built a very good bridge between linear model theory and designed experiment. Just imagine if you only have a calculator, not even the scientific one, and some scratch paper, can you finish analyzing a design problem? This is where they catch up, in traditional designed class, you learn the design, you learn the model, the ANOVA table, the test statistics, and many formulas. In a linear model class, you learn the general estimation, you learn the test, the asymptotic distribution. But I bet every time the calculation is left to the computer. So here is the question, what if some assumptions are violated? Can you help me solve it? What if I am not satisfied with Satterthewaite, can I go to the data, the box plot, and find out which sets of EU'S are making estimation go counter intuitive? These are the question we need to answer in the real life, textbook homework won't ask you this. And yes, this textbook will guide you through the steps needed, no matter methods of moment or REML, God is always in the details!But I still recommend having a textbook on linear model theory like plan answers by Christensen, it can make you feel that statistics is just simple calculation, nothing mysterious.

Textbook for Experimental Design for Product Development. So far I have only used it as a complimentary resource to learning 2^n factorial designs, but it is well written. Also helps that one of the authors is my professor who explains statistical concepts well for applications and not pure theory. I always feel that I understand these concepts better when they can be placed in a 'real world setting'.

This book has wonderful explanations that are easy to follow and several examples to better understand the material. It's a great resource that I have referenced on several occasions. I highly recommend the text.