Workshop Enrollment: Logistic Regression for Binary, Ordinal, and Multinomial Outcomes

A 4-Module Online, Interactive Workshop

Starts on Thursday, April 25, 2019

Sooner or later, you’re going to have to answer a research question with a categorical dependent variable:

  • Are bilingual speakers more likely to make pronunciation mistakes than native speakers under conditions of cognitive load?
  • Did a physical therapy intervention improve the likelihood that athletes with a knee injury were able to return to play within the same season?
  • Do nutrition protocol and a physical therapy intervention affect whether patients can perform each of a set of tasks post-surgery?
  • As you may have already encountered, no matter how many ways you transform or try to finagle the data, you just can’t force it into a linear regression or ANOVA.

So what do you do?

Logistic regression: A researcher’s best friend when it comes to categorical outcome variables.

Maybe you’ve avoided logistic regression before because it’s seemed quite complex or overwhelming… or simply because it wasn’t a required part of your previous statistics coursework.

It’s time to get you over that barrier.

Yes, understanding logistic regression will require some new statistical concepts, but we assure you:

If you can use linear models, you can understand logistic regression (really!).

I’m Karen Grace-Martin, workshop instructor.

My goal is that by the end of the workshop, you will be able to recognize the need for, conduct, evaluate, and interpret a logistic regression model, as well as use it to predict outcome probabilities.

In this four-module live, online workshop, you’ll learn the similarities and differences between linear and logistic regression for three kinds of categorical variables:

  • Binary (two categories)
  • Multinomial (three or more unordered categories)
  • Ordinal (three or more ordered categories)

All those concepts unique to logistic regression, like information criteria, maximum likelihood, deviance, logit links, odds ratios, and ROC curves? We’ll go over those in detail.

After this course, you will fear no categorical dependent variable!

Karen Grace-MartinKaren Grace-Martin MA, MA

Who Is This Workshop For?

This workshop is for you if you:

  • You’ve attempted a logistic regression before, but found it confusing or difficult, and you didn’t really understand it
  • You have used linear regression or ANOVA and now want to expand your knowledge
  • You know you’ll soon be faced with the need to implement a logistic regression and want to know how to proceed

It is NOT for you if you:

  • You’re new to statistics or regression and ANOVA (You’ll need a solid background in statistics. See below for more details on the course prerequisites.)
  • You don’t have time to keep up with the coursework. (To participate fully, please be prepared to invest 5-10 hours per week for workshop sessions, Q&A, and exercises. While you can revisit the material any time during the next 12 months, if you want to take full advantage of the live components, set aside time NOW.)

“I’m so glad I came across [this] site and registered for this workshop. I’m sure I’ll be registering for more workshops in the future! The cost-benefit for this workshop is incredible! You just can’t beat it!”

— Paula Fleshman, PhD, Education Research Consultant, Previous workshop participant —

How Does It Work?

This course is a 4-module live, interactive online workshop.

All lessons are conducted online via live webinar. You’ll see the instructor’s screen to view the presentation…all from your own home or office.

During each webinar session, the instructor will cover core concepts and steps and will leave plenty of time to ask your own questions.

Have more questions? No problem!

In addition to the four lessons, you’ll also meet with your instructor for four additional Q&A sessions where you can get additional assistance on workshop concepts, deepen your knowledge, and clarify any questions.

All 4 Workshop Sessions and 4 Q&A Sessions Are Recorded for Your Convenience.

Where Do I Access Recordings and Material?

For each of the four modules, you will have a set of training resources and material.

As a participant of the Logistic Regression workshop, you can access these for 12 months in our participant-only website.

Data Sets
Real research data sets in SPSS, SAS, Stata, and csv formats.

Exercises + Solutions
Try out what you’ve learned in each module to cement the material. Never get stuck with provided syntax and interpretations.

Syntax and Demo Videos
Step-by-step videos for SPSS, SAS, Stata, and R. Available to download so you can save and edit them for your own analysis.

Forum Access
Ask your instructor about workshop content or how to apply it to your own data analysis.

Handouts
Session slides and supplementary material to support your learning.

Q&A Recordings
Can’t make a live Q&A? Submit questions ahead of time and get answers from the recording.

Session Recordings
Made available within 48 hours so you can review material at your preferred schedule.

Suggested Resources
External videos, articles, or exercises that we recommend to support your learning.

You’ll have access to this site and all the related materials and resources for ONE FULL YEAR. That means you can re-watch sessions and ask additional questions again during that 12-month period.

Often, our students report they understand the material at a deeper level on a second or third pass. You’ll learn more every time, so take advantage of it!

“…You really get the method at a fundamental level, but not in a way disconnected from the grit of actual practice… Karen strikes the perfect balance of theory and practice.”

— Fred Petillo, MS, MBA, Market Research Manager, State Bar of Wisconsin, Previous workshop participant —

What’s Covered in the Workshop?

Module 1: The Binary Logistic Regression Model

This module lays the foundation for logistic regression. We’ll start with a review of linear models, then walk through differences and similarities to logistic models. We’ll talk in detail about why the logistic model is necessary and what a logit link really means. Next, we’ll review probabilities and odds and what they really measure. Then we’ll end with a simple example, and walk through what the model tells us about the relationships between predictors and our categorical outcome.

We specifically will cover:

  • A brief review of the linear regression model, why it does not work for categorical data, and why logistic regression does
  • The similarities and differences between the linear and logistic models
  • Link functions, the logit function, probability and odds, and how they are used in logistic regression
  • Model assumptions
  • Interpreting coefficients, odds ratios, relative risks, and confidence intervals, including odds ratios for categorical predictors and interactions
  • Demonstrations in SAS, R, Stata and SPSS (pre-recorded)

Module 2: Fitting and Evaluating the Model

Now that we’ve got a solid understanding of the model, we’ll learn ways of evaluating it. Is it any good? Does it fit the data? How well does the model predict outcomes? What happened to R-squared?

  • Maximum Likelihood estimation: what it means and how it affects model fit statistics
  • Evaluating the model as a whole using Likelihood Ratio Tests, Deviance, Information Criteria, and Pseudo R-Squared statistics
  • Using the model to predict probabilities and classify individuals with classification tables
  • Evaluating the predictive ability of the model using ROC Curves and Hosmer-Lemeshow Goodness of Fit tests
  • Demonstrations in SAS, R, Stata and SPSS (pre-recorded)

Module 3: Using and Assessing the Model to Predict Outcomes

Once we know a model as a whole is working, we can then investigate in detail the effects of individual predictors. There are various ways to do so: graphing, calculating predicted probabilities and examining odds ratios. Odds ratios in particular are bizarre, but you can learn them. We’ll spend a lot of time on interpreting them, especially for interactions, which get even trickier.

  • Evaluating each predictor using Likelihood Ratios and Wald Statistics
  • A step-by-step of how to use the estimated coefficients to calculate predicted probabilities and use them to interpret effects of predictors
  • Interpreting odds ratios for interaction effects
  • Model assumptions and dealing with data fitting challenges, including complete separation, zero cell counts, and influential points
  • Demonstrations in SAS, R, Stata and SPSS (pre-recorded)

Module 4: Ordinal and Multinomial Models

Everything we’ve done up to this point has been about binary models, as they’re the foundation of logistic regression. Good news: Everything we do in binary logistic models can be expanded to multinomial and ordinal categorical outcomes, so that’s our focus for this module.

  • Differences and similarities between the multinomial model and the binomial model
  • The different procedures in SAS, R, SPSS and Stata you need to run multinomial models
  • The difference between ordinal models for 3 or more ordered categories and multinomial models for 3 or more unordered categories
  • The big effect that software default settings can have on your results
  • The proportional odds assumption
  • Demonstrations in SAS, R, Stata and SPSS (pre-recorded)

Dates and Times

There are four workshop modules, each with a lesson and a Q&A session (plus supporting materials).

We will meet on Tuesdays at 3pm, US Eastern. The workshop sessions are approximately 2 hours long, including time for questions.

Workshop Orientation (Optional):

Wednesday, April 24, 2019 from 3pm – 4pm (GMT-5)

Live Workshop Dates: Thursdays from 3pm – 5pm (GMT-5)

Module 1 – April 25

Module 2 – May 2

Module 3 – May 9

Module 4 – May 16

Q&A Session Dates: Wednesdays from 3pm – 4pm (GMT-5)

Module 1 – May 1

Module 2 – May 8

Module 3 – May 15

Module 4 – May 22

Remember, everything is recorded and available for you to watch at your convenience, should you be unable to attend a live session.

Also, you have access to all the class materials for a full 12 months from your enrollment date.



Sorry, enrollment for this workshop has closed.

Please contact us if you have any questions
or get on the waiting list to be notified when it's offered again.


About Your Instructor

Hi, I’m Karen Grace-Martin, your workshop instructor.

As president and founder of The Analysis Factor, I’ve been supporting researchers like you through their statistical planning, analysis, and interpretation since 1997.

My goal is to help you improve your statistical literacy so you can bring your important research results into the light with confidence.

Karen Grace-Martin
Karen Grace-Martin MA, MA

“Karen is a brilliant presenter. She has a very good command of the subject matter and is able to present complicated material in a way that is easy to understand, while at the same time conducting the workshops in a professional manner.”

— Philip O’Sullivan, Senior Policy Advisor, Insurance Council of Australia, Previous workshop participant —

Prerequisites

So what kind of background in statistics do you need?

This workshop is for researchers, not statisticians. That being said, it is an INTERMEDIATE LEVEL workshop.

You should have solid experience running linear models (ANOVA and/or linear regression), including a basic understanding of:

  • Least-squares estimation
  • Dummy variables
  • Interactions

Be prepared to do a bit of algebra and equation solving.

You’ll get the most from this workshop if you have:

  • MINIMUM two statistics classes, including some experience in data analysis, particularly linear modeling.
  • Familiarity with using a statistical software package. Examples, data sets, and code will be provided for SPSS, SAS, Stata and R.

An additional word about software packages:

This workshop focuses on the concepts, steps, and interpretation of logistic regression — it is not about the software. While I am familiar with Minitab and JMP and will do my best to answer software questions, I’m not familiar with the intricacies and defaults of every software program out there. However, if YOU are familiar with another package and are comfortable reading the manual to understand the defaults in its logistic regression procedure, you’ll be fine.

If you have questions about whether you’re ready, just email us. We’ll give you our honest opinion. We want you to succeed!

Your Satisfaction Is Guaranteed

As with all our programs, your satisfaction is guaranteed. If you participate fully in a workshop–watch, read, and try out everything included–and find you are not satisfied for any reason, we will give you a full refund, no questions asked. Just notify us within 90 days of purchasing the program.

FAQs

Q: I haven’t taken a statistics course in years. Will I be able to keep up?

A: As long as you have the suggested prerequisites (a strong understanding of linear regression), you should be fine.

However, you WILL need to spend some time each week, both going over the material and also “getting your hands dirty” with some data. Expect to spend 4-5 hours per week reviewing the material and doing the exercises, and re-running class examples on your own.

And if you have questions during the workshop, make sure to attend the Q&A sessions and/or post on the Workshop Center to clarify anything you’re not clear on.

You know how this goes: The more time you put in, the deeper your understanding.

Q: What’s your refund policy?

A: Your registration fee is fully refundable up to 72 hours before your first class session. No refunds will be granted after the program begins.

Q: Do you offer student discounts?

A: Yes, we do! Current students with a valid ID can register for up to 20% off the standard rate. Click here to receive our student discount on all our workshops: Student Discount List

Q: Things are very busy at work right now, and I’m afraid I won’t have time to keep up. How much time per week does this really take?

A: We’ve worked hard to make sure people can take a workshop in the midst of their normal work life. For longer workshops, we build in a week off halfway through the course.

Each workshop webinar session is about 2 hours, whether you attend live or listen to the recording.

Each week, we have a one-hour Q&A session. You don’t have to attend, we just make those available so you can have any questions answered or listen in on others’ questions.

We provide exercises, answers and data sets, so you can practice what you’ve learned each week and get help in the Q&As. These will take you a few hours each week, maybe up to 3 or so. The exercises aren’t required—you’re not graded. But we’ve gotten feedback that they’re very helpful, as it allows you to practice and figure out what you did and didn’t understand.

In other words, each week, there are approximately 2 hours of instruction, one hour of Q&A, and about 2-3 hours of practice time. So if you have a crazy hectic week during the workshop, you have catch up time. But If you have a crazy hectic month, you may be better off waiting as you won’t really have a chance to catch up.

One reminder: You have access to the workshop website, all materials, and a place to ask questions for a year. So if you get behind, you aren’t missing out. You can catch up on your schedule.

Q: I teach/work/sleep during this time and cannot attend live. Is there any way to register and access the webinar sessions later?

A: Yes! We have participants from many time zones with many different work schedules. In order to support our diverse student base, all material – including the workshop sessions and Q&As — are recorded and made available to students within 48 hours. They’re screenshot video files, so you’ll hear the instructor talking and see his screen, just as live participants do. You can even submit questions for written answers any time between sessions. Many students take the entire workshop in this manner, never attending a session live.

You can also download and keep these videos forever, so if you need to refresh your memory at any point in the future, it’s yours to review.

Q: I’m outside the U.S. Can I still participate?

A: Yes. We have participants in our workshops from many different countries. You will want a fast internet connection and either a computer speaker or a telephone if you plan to attend live.

Q: Can I pay with Paypal?

A: Yes. When you check out, you can pay directly using our system or paying with Paypal. However, the three-payment option is not available using PayPal.

Q: Can I join the Q&A webinars from my tablet or phone?

A: Yes. You can join all webinar sessions from your Apple or Android device by downloading a free GoToMeeting app.

Q: This really isn’t a good time for me. Will you be offering this workshop again soon?

A: We don’t have a set schedule for running each workshop. We try to schedule each workshop every 12-18 months, though very popular workshops may be offered more often. We are sometimes unable to re-run a workshop if the instructor becomes unavailable so there are no guarantees that any workshop will be run again.

If you think you’ll need the material in the next half-year or so, I would suggest registering now. You’ll get all the materials to download and review at your convenience, and you can attend again for free any time in the next 12 months.

Have additional questions? We’re here to help! Just email us at support@theanalysisfactor.com.