# What are the knowledge/books needed before learning the Multivariate Data Analysis (Hair et al.)?

Multivariate Data Analysis (Hair et al.) is the most popular book cited in marketing papers. However, I found it hard for me to understand. I have learned basic marketing research in my undergraduate course (like using SPSS to do t-test, ANOVA, and regression). I don't have other mathematics training.

Now I'm learning this book by myself for a thesis. I can't find any video courses online using this book. I can only read this book bit by bit, and I'm reading it very slowly (three days for half a chapter). The mainly painful issue is many things I don't understand and I often become distracted.

So what are the knowledge/books needed before learning the Multivariate Data Analysis book by Hair et al? The most useful parts for me are MANOVA and SEM. Is there anything I can skip in this book, or any order I should follow (my experience is I started with MANOVA, and found it discussed a lot on discriminant analysis, so I went back to read discriminant analysis...but still feel confused sometimes...)?

• Maybe first a course in calculus, or maybe directly linear algebra? Aug 14, 2017 at 1:20
• Before diving into Multivariate statistical methods you should have a solid understanding of univariate statistical methods such as linear regression, ANOVA, t-tests, etc. Multivariate statistics deals a lot with linear algebra, so a course or two (I took two) before diving into the material would be helpful.
– Jon
Aug 14, 2017 at 16:06

## 1 Answer

I flagged this question as primarily opinion-based (and I still think it is) and it was declined. So, I will try to provide an answer since I am familiar with Hair et. al. (2014) Multivariate Data Analysis [international edition, it looks like this is an important information].

Although a first course in calculus or linear algebra is always useful as kjetil b halvorsen mentioned in his comment (and required if you want to have a better grasp of these methods), I don't think you need to take a such a course to understand this book. I think the language of the book is relatively easy, especially if you are not familiar with the mathematical background. So, if you experience difficulties, it might be better to refresh your memory.

For instance, before reading the MANOVA and GLM chapter, you might want to review ANOVA. Field, Miles & Field (2012) Discovering Statistics Using R (there is also an SPSS version, if you are more comfortable with it) provide 5 chapters on ANOVA and related methods (GLM 1-5) as well as MANOVA. It is an introductory textbook and you can find many examples that you can replicate using the related software.

You can follow a similar strategy for the SEM chapter. So, reading on factor analysis and multiple regression might be helpful. This means you might want to start with earlier chapters on exploratory factor analysis and multiple regression. Please note that the chapter on confirmatory factor analysis follows the chapter on structural equation modeling overview. This is for a good reason as it requires some understanding of SEM. You might also want to read about path analysis which is covered in SEM overview chapter, but not separately in Hair et. al.'s book. Still, I am not sure that Hair et. al.'s book is a good place to start learning SEM (if that is what you are doing). I recommend Kline's (2015) Principles and Practice of Structural Equation Modeling.

Although the answer is inevitably opinion-based, I hope it would be helpful.