You have spent months designing and creating your study, and collecting data. Your data are entered, you start to analyze, and your statistical software drops half of the hard-won data due to missing data. Now there isn't enough power to find any significant results, even though the trends are there.

A colleague told you to just impute the mean to fill in those missing data. But now the paper has been rejected--the reviewer says mean imputation gives biased results (which it does, by the way). Now what do you do?

This workshop will get you up to speed on the modern approaches to missing data that give you unbiased results with no loss of power--multiple imputation and full information maximum likelihood.

You will learn what they are, how to implement them in statistical software, when they are needed and can be used, and how to check their assumptions.

You will also learn about the old, simple techniques and how they solve problems with missing data, the new problems they create, and the situations in which they work well enough.

Session 1:

The three types of missing data--what they are, why they matter, how they affect your analyses, and the steps to figure out which one you have.

The common traditional, simple approaches--listwise deletion and imputation--when they do work, when they really, really don't, and why.

How to impute using the EM Algorithm--a simple, unbiased way to impute data when only small percentages of data are missing.

Two new and drastically better approaches--multiple imputation and full information maximum likelihood--why they're better, when you can use them, how to do them.

Session 2:

Mulitple Imputation: what it is, when you can use it, how to do it step-by-step

Session 3:

Full Information Maximum Likelihood: what it is, when you can use it, how to do it step-by-step

Bonus Session:

Two weeks later, we will meet for a question-and-answer only session. Use those two weeks to try out what you've learned, then bring your questions, either general or specific to your data analysis.

• May 21
• May 28
• June 4

All classes meet on from 11:30 am-1pm eastern.

• Have struggled with the devastating loss of power that comes from missing data
• Realize that listwise deletion and mean imputation don't usually work well, and are looking for a better way
• Have heard about the amazing miracle of multiple imputation and want to learn what it is and how to do it
• Have struggled with using multiple imputation and realize that it can be quite difficult to implement well. You want to know when is it really necessary, and when (and how) can you use Maximimum Likelihood instead, which is both simple and powerful

Prerequisites:

• You will get the most out of the workshop if you have had at least two statistics classes and experience in data analysis.
• You use SAS or SPSS. You are welcome to use another software package, but I am only familiar with using Missing Data procedures in these software packages. I know that R, S-Plus, and Stata have multiple imputation capabilities, and I have used these packages before, so I'm mildly familiar with them, but I don't know the specifics. Note: SPSS can only do multiple imputation in version 17.0, but there is a work-around for earlier versions, which I will show you. For any of the SPSS work, you will need to have the missing values add-on module. If you have it, "Missing Values" will appear in your Analyze menu. If you don't and are employed by a university, you can get a one-year license for Windows or Mac to the full SPSS suite, including all their modules at On the Hub. Note that the Grad Pack does NOT contain the Missing Values Module, but the Faculty Pack does.
• We will use AMOS for the Full Information Maximum Likelihood. AMOS now comes bundled with SPSS. No prior experience using AMOS is necessary. Full Information Maximum Liklihood can also be run in MPlus, but I do not know how to use it.

• 3 - 90 minute workshop sessions.  Each session will contain a presentation and time for questions.  
  
• A 60 minute Q&A session June 30 at 11:30 am. Just your questions and my answers. This is two weeks later, so you have a chance to try things out on your own data, then come back with questions. It's when you use these techniques on your own data that you cement your learning.
• Data files and SAS and SPSS code to run and explore all of the examples yourself.
• Video Recordings of each workshop session made available within 48 hours. You can watch these again and again as new issues come up in your analysis.
• A workshop message board where you can ask questions between sessions.
• Participants are encouraged to ask questions about implementing these techniques in their own analyses in addition to general questions about the topic.  (But please keep questions to the topic of the workshop so that they’re useful for everyone). 
• Homework.  (Yes, homework!).  It’s optional, but you’ll get more out of the workshop if you participate fully by doing homework.  You can use one of the data sets provided, or use your own data.

Your registration fee is fully refundable up to 72 hours in advance minus a $25 administration fee. Because enrollment is limited, no refunds will be granted after the program begins.

That said, your satisfaction is guaranteed. If you participate in the full workshop and find you are not satisfied for any reason, we will give you a full refund. Just notify us within 90 days of the conclusion of the program.