Navigating the Complex World of

SPSS Training

I have taught all of the SPSS Modeler courses and SPSS Statistics courses many many times, and I would be glad to give you advice regarding any of them. I hope that these extensive course notes help you find the class that you need. I will try to maintain a discussion of every available class in this guide. If you find it a little overwhelming, and many have, just shoot me an email and I will try my best to help.

If you want to take an ‘official’ class I can help you choose a Global Training Provider (GTP) offering that meets your needs, and I might even turn out to be your instructor. If I’m scheduled for an upcoming event, and there is an IBM GTP course that first your needs I can let you know my schedule.

If you want to arrange a custom class at your location or online using your data in lieu of the official course guides I would be glad to help with that as well. The online option means that I might be able to help you no matter where in the world you might be located. Those are not the only two options. Feel free to reach out to me with questions. One of the advantages of a custom class is that you can always use your version - Version 17, or whatever your version happens to be. Another advantage of the custom approach is that you can combine bits of pieces from several classes into one custom class. I've dedicated a recent post to what I consider to be the best strategy for SPSS Modeler classes.

In recent years IBM has decided to involve five trusted Global Training Providers (in 2013 there were only 4) to conduct their training: Steve Mills on IBM Training: Building skills for a smarter planet. They all do a great job, and they have the very important role of alerting the public to what is available and matching those that need training with those that can provide it. You may have a relationship with one that allows you to take a variety of IBM classes. They each have unique pro and cons, and differing business models.

For the moment, IBM continues to provide the content, but there are indications that the partners will be encouraged to tailor, customize, and regionalize the content. Ask your GTP what work they are doing in this area. Those GTPs that are doing the best work in keeping the content up to date deserve your training dollars. The GTPs primary roles are to provide classrooms, logistical support, customer service, and find trainers.

However, there are a limited number of SPSS expert trainers out there, so they all need to draw from the same small pool of SPSS talent. Don’t be afraid to ask who your trainer will be. Ask them if they have 5, or 10, 20 or even more than 20 years of experience. Ask them how many times they have taught the class. That should be one of the factors you consider when choosing a GTP and a class.

Below are my comments and advice regarding the SPSS Modeler and SPSS Statistics curriculum with several topics and subtopics. To view them, simply click on the chapter to expand the content.

IBM SPSS Modeler Courses

 

Introduction to IBM SPSS Modeler and Data Mining

This is designed as a first course. Note IBM's course map, but I'll walk you through all of the courses here. IBM’s GTPs offer this in a two day format, and three day format. The more recent version 0A005  is for Version 16, and is two days. I’ve always felt the three day version is probably better for most students. Most of the content that has been reduced to fit into two days are found in the modeling chapters, which for many is the most important content. There are other ways to reduce the material to fit into two days. The older book 0A002 is better in some ways, but is written for Version 15. The current version is 17. A custom class using your data is always an option, and for virtually all a better option. However, a public class is the most cost effective.

Understanding Self Paced Virtual Classroom

There are also to self-paced courses. 0E003 is a two day format of the Intro for Version 14. 0E002 is a three day format of the Intro. Almost every course is available in this format. It is important to understand what Self Paced Virtual Classroom is. A major advantage is that you get access to a learning environment allows you to use Modeler remotely. This is great, if you don’t own Modeler. Now that there you can get a trial, this might not be as big an advantage. If you own your own copy of SPSS Modeler (or have access to a trial), however, there is no other sense in which there is a ‘classroom’. It is basically a book, data, and practice streams (streams are Modeler’s code). There is no asynchronous instruction, per se. There are no videos to watch or software demonstrations to watch. You must be prepared to read a book, and it is hundreds of pages. You may find it difficult to complete the book within the 30 days. It may be your only option in this price range, but be prepared to commit the necessary hours of reading and practicing on your own.

Advanced Data Preparation Using IBM SPSS Modeler

This one day course may not be quite what one expects. It includes some good content, but it is not all ‘advanced’ content. These topics include some topics that had to be dropped from the Introduction course as the format has been lengthened and shortened over the years. When an expert instructor knows which application areas you are most interested in, they can usually cover the most important material in 1/2 a day. In this way, it is not difficult to include this material in a custom three day intro course. The codes for this course vary by software version and format: 0A055G, 0A052, and 0A054.

Modeling Courses

This is a big and important topic, especially since the Introduction course does not cover this topic extensively. It is also very confusing because there are two different versions of the Modeling curriculum. The ‘old format’ had just three classes, and the ‘new format’ has several classes. In may look at first glance, that there are 7 different modeling classes, but it isn’t really true. My explanation of the two versions of the curriculum should make this clear.

‘Old Curriculum’ Modeling classes

Predictive Modeling with IBM SPSS Modeler 0A032

This is a great class, three days in length, but unfortunately it has been largely discontinued. It is great to see many techniques compared and contrasted with each other. It covers all of the algorithms that predict either categorical or continuous targets. Since it was three days in length, it had enough time to cover supporting topics as well. The new format course break this course into multiple one day courses.

Clustering and Association Models with IBM SPSS Modeler 0A042

Since these techniques, K-Means clustering, Kohonen Networks, Association Rules, etc., were not ‘predictive’ techniques, they got their own one day class. The name has changed slightly, but this course remains largely unchanged in the new format.

Text Analytics has always been taught in a two day format. The format has not changed, but an updated version exists for the new format curriculum.

‘New Curriculum’ Modeling classes

The following three courses are one day courses that replace the old format three day Predictive Modeling with IBM SPSS Modeler course:

Classifying Customers Using IBM SPSS Modeler 0A0V4 - Includes all of the classic techniques including Decision Trees. This is, for most users, the critical follow on after the Intro course unless they find the three day version on offer.

Predicting Continuous Targets Using IBM SPSS Modeler 0A0V4 - as the name indicates these techniques apply when the target variable is continuous. Famous examples include Linear Regression and Regression Trees.

Advanced Predictive Modeling Using IBM SPSS Modeler 0A0V4 - Not really ‘advanced’ per se. This could also be called ‘machine learning’ modeling. These are simply the techniques that are less statistical. If you already knew the statistically based techniques you might choose to take the first. Perhaps surprisingly, you might also take this one first if you know nothing about the statistically based techniques because these techniques assume little knowledge of statistics.

Much like the old curriculum the next two courses supplement the predictive modeling.

Clustering and Association Models Using IBM SPSS Modeler 0A044 - This simply replaces and updates the Clustering ‘with’ course for newer versions. Taken with the other three, the total of four days of standard modeling instruction.

Introduction to IBM SPSS Modeler Text Analytics 0A104 - Not yet updated for version 16, this course has remained largely unchanged. If you own Modeler Premium you will want to orient your self to these powerful techniques.


IBM SPSS Statistics Courses

How to Get Started: The Core SPSS Statistics Curriculum

It is important to remember the SPSS Statistics is built around modules. When you begin you want to start by mastering SPSS Base. The current version is Version 23. The official materials may lag a version or more than one version behind. Note IBM's course map, which will list the versions. Make sure to ask. I'll explain all of the courses on this page. The first two courses should be:

Introduction to IBM SPSS Statistics 0G503 &

Data Management and Manipulation with IBM SPSS Statistics 0G529.

The first course assumes nothing at the start, and is always a good foundation. If you are afraid it would be too basic consider a custom class, and accelerate a bit. In a custom format these first two classes can easily be covered in three days if everyone in class works on the same team. Data Management and Manipulation was called ‘Intermediate’ years ago, and has always been designed as a class that nearly everyone should take. Although it gets into some topics that some might consider optional (like SPSS Syntax), it covers them briskly. And there are some topics that everyone must know like Merge and Aggregate.

The Self-Paced Option for SPSS Statistics

These same two introductory courses are available as a Self Paced Virtual Classroom, currently coded 0K502 and 0K522.

Keep in mind that part of what you are paying for are the “labs”, and big part of the lab cost is a remote copy of SPSS Statistics that you can use if you don’t have your own copy. This might be valuable to you, but if you have your own copy you might not need this. Even if you do not own it, you can get a trial copy. The SPVC option includes no videos or software demonstrations. No live instructor. You are basically reading a book. There are some practice data sets, but SPSS ships with plenty of practice data sets and case studies. They come with the software. SPVC also offers some email support. Don’t forget the most important detail - at the end of 30 days you lose access to everything including the book and any work that you have done in the remote environment.

They are extremely popular, but you want to make sure that you are prepared to take the lead on your learning. I think the reason they are popular is that unlike live instruction, they don’t get cancelled, but they are somewhat cheaper than live instruction. In the case of SPSS Statistics, in particular, there are tons of alternatives, however. There are dozens of third party books on Amazon including SPSS Statistics for Dummies, 3rd Edition, co-authored by myself. A book like this won’t cover everything that is covered in the first two courses, and it is also just a book, but it will get you started. Make sure to explore your options as taking the first two courses in this format will cost several hundreds dollars each. That is less expensive than live instruction, but it is expensive compared to a book. You may want to consider buying a less expensive book, and purchasing a few hours with an expert to answer questions with the money that you will save. And then you get to keep the book after 30 days! Don’t forget to verify version. A custom class can always be taught in your current version.

Going beyond the basics with SPSS Programming

There are three classes on the programming theme:

IBM SPSS Statistics Syntax I 0G406

IBM SPSS Statistics Syntax II 0G416

Integrating Functions, Procedures, and Algorithms with IBM SPSS Statistics 0G437

All three will assume knowledge at the level of the Data Management and Manipulation with SPSS Statistics course. Years ago, I helped write some of this material. That course guide has been retired, but was an intermediate level treatment of Syntax. A great way to learn this material is to apply it to your situation. The third course gets into some of the more recent functionality in SPSS Statistics programming often referred to as ‘programmability’.

Going beyond the basics by reviewing some Statistical Theory

This is a big topic, but you can learn much of what SPSS Statistics Base can do by taking the following three classes. Although some mention of techniques that require the modules will be made (especially in ANOVA) these are consider foundation course. Over the years, I’ve had many SPSS training clients take all three of these with me, sometimes all in a row. As always, ask about software version, although that may be less important for these classes as the theory does not change, and change to the functionality of these menus does not occur with every version. If you want to ask me about your specific situation regarding this, don’t hesitate to email me.

Introduction to Statistical Analysis Using IBM SPSS Statistics 0G512

The course assumes very little theory, but does assume that you have taken Introduction to IBM SPSS Statistics. The are completely different courses. “Stats Analysis” is a theory course covering basic foundation topics like hypothesis testing, Chi-Square, T-Test, and Simple Linear Regression. You will learn the theory, the point and click steps to get results, and the interpretation of the resulting output. Always been one of my favorite classes.

Advanced Techniques: Regression 0G073

Another one of my favorites, this one is not ‘advanced’ in the sense that you might think. It is called that because it requires that you know the material in Introduction to Statistical Analysis, but it starts from the very beginning. So it is a basic introduction to Simple, Multiple, and Logistic Regression. By the end of three days, however, you have covered a lot of ground. At about 20 classroom hours, it is about as much depth as a course in Regression in a university setting - at least I try to duplicate that much coverage when I’ve taught it.

Advanced Techniques: ANOVA 0G063

Most participants find this one a little more intense. If all you need is T-Test and One-Way ANOVA then Introduction to IBM SPSS Statistics might be enough. This course gets into more advanced topics like Multi-Way ANOVA and MANOVA. It is also covers HLM and a Split Plot. If you don’t recognize these terms then research this one more before you sign up. If you have questions, don’t hesitate to email me. Some of the procedures will requires the IBM SPSS Advanced Statistics module.

More Theory Classes for SPSS Statistics

These three classes are not frequently found on the GTP schedules, but include some useful content.

Survey Analysis Using IBM SPSS Statistics - ILT (0G317)

The Survey Analysis class includes virtually all of the techniques is the three days Advanced Stats class. As such, it is an intense experience. What makes it different is that it focuses on likert type five point scales and other non-scale variables.

Market Segmentation Using IBM SPSS Statistics - ILT (0G036)

The Marketing Segmentation class has always been a favorite of mine. It basically covers unsupervised segmentation aka Clustering using technique like K-means, and supervised segmentation using techniques like Decision Trees. Note well that while some of these techniques can be done with SPSS Base, some others require modules, notably Decision Trees.

Advanced Statistical Methods for Healthcare Research - ILT (0G016)

The course is a reworking of the Advanced Statistical Methods course using healthcare data and example. It is rare on public schedule, but custom training is always an option, and can be arranged with your own healthcare data.

The SPSS Statistics modules

Each of these classes is dedicated to a different SPSS Module. The module system can be confusing, but the purpose is to prevent the pricing of SPSS Base to get out of reach of customers. By adding functionality a la carte, it keep the price of SPSS Base in check. Before taking one of these classes publicly verify that you will be able to use the features when you get back to the office. In other words, verify that you have a current license for the relevant module.

Introduction to IBM SPSS Neural Networks - ILT (0G046)

Introduction to IBM SPSS Decision Trees (V19) (0G2K9)

Introduction to IBM SPSS Complex Samples - ILT (0G023)

Correspondence Analysis and Multidimensional Scaling w/ IBM SPSS Categories -ILT (0G056)

Time Series Analysis and Forecasting with IBM SPSS Forecasting - ILT (0G096)

SPSS Statistics companion products

These two products, and their courses, cause a lot of confusion.

Introduction to Structural Equation Modeling Using IBM SPSS Amos (V22) (0G203)

AMOS is a standalone product that was originally made by a company called smallwaters. SPSS Inc. acquired the product, which then in turn, became part of the IBM SPSS family of products. It has its own interface, and is used to perform Structural Equation Modeling (SEM).

Introduction to IBM SPSS Text Analytics for Surveys - ILT (0G300)

SPSS Text Analytics for Surveys (you will sometimes see the old acronym STAfS) is also a standalone product with its own interface. Naturally, there is frequent confusion with the equivalent product within SPSS Modeler. There have some of the same underlying technology, but the interfaces are very different.

I enjoy teaching both of these products, so for more information, or to make a training inquiry, or for consulting contact me.