Thank you to everyone who filled out my survey last week about which workshops you’d like for Winter/Spring 2010. It was very helpful. I wanted to lay out the results here, rather than in an email, so that you could continue to give me your feedback in the comments.
There was a lot of interest in all the workshop topics, but the clear winner was “Interpreting Regression Coefficients: Dummy Coding, Interactions, Centering.” 69% of respondents replied “Yes, definitely interested” and it had the most votes for the information needed most urgently.
So we’ll schedule that one to begin in January. If you find you need the information right now, we do have available the transcripts from the last time we offered the workshop, in ebook form. They’re edited to read nicely and each section begins with the corresponding presentation slide, so you have the visuals. It’s not the same as being part of a workshop, but not everyone needs or can attend a workshop, and they are available right now.
A close second was “Dealing with Multicollinearity” with 62% intense interest, followed by Logistic Regression and Missing Data, with 58% and 51%, respectively. I do hope to offer all of these over the next year, so keep your eyes peeled for announcements. In the meantime, our next webinar is on defining and diagnosing Multicollinearity. We can’t fit solutions into an hour in a useful way, but there is a lot of misinformation about diagnosis out there, so this will get you started.
Mixed Models for Repeated Measures
The most interesting part of the survey, though, were the write-in results. Two topics with multiple requests were Structural Equation Modeling and Factor Analysis. I hadn’t considered either one any time soon, but I will start thinking about them.
Even more people were interested in a workshop on running repeated measures data as mixed models. It’s a really important topic, and it’s not easy, but becoming more and more necessary. One idea I had was to include a consulting program option along with the workshop. In small groups (maybe < 6-8 people), we would simultaneously run a series of consulting sessions, where we go over your model–how to run it and what it means. It would cost much less than individual consulting, but give more support and guidance than a workshop.
One nice thing about small groups is you get the opportunity to see many examples. It’s the best way to really understand modeling.
Then it struck me that I could do this with any of the workshops.
So leave a comment to tell me:
1. Which of these, or other, topics do you want to see covered in the winter/spring?
2. Would you like the program option with the mixed model workshop?
3. Would you like the program option with other workshops? If so, how should we set it up? What features would you like in the program?