June 4-6 2001

1=disagree 5=agree

• The course met the objectives advertised on Insightful's websight and flyers
  • 5: 5/9 4: 4/9
• has provided me with a good foundation for working with S-PLUS
  • 5: 5/9 4: 4/9
• The instructor demonstrated a thorough knowledge of the course material
  • 5: 9/9
• fostered a good learning environment
  • 5: 8/9 4: 1/9
• was well prepared and organized
  • 5: 9/9
• What are the best aspects of this course?
  • The examples, hands on and learning many new commands written by F. Harrell
  • The instructor, material and your product. Frank is the best I've ever seen.
  • Hands on experience. Instructor's experience with data and his philosophy of data analysis
  • Frank is incredibly knowledgeable and well organized. Very good course materials.
  • Combination of lecture and lab forms; very clear presentation and examples; instructor shows extraordinary knowledge and perception in subject area.
  • 1) Shrinkage; 2) Missing value imputation
• Suggestions for future:
  • Send out some course material in advance
  • More integration of lecture and examples. Show short examples with code as the notes are followed through.
  • Hands-on exercises were only useful as a change of pace. I found it more useful to simply watch or listen to examples rather than trying to keep up with typing.
  • Lab work seemed rushed sometimes - leave more in-class or free time for student computer work.

June 22-24 2002 Executrain, Connecticut Ave., DC

1=disagree 5=agree

• The course met the objectives advertised on Insightful's websight and flyers
  • 5: 7/9 4: 2/9
• has provided me with a good foundation for working with S-PLUS
  • 5: 4/9 4: 5/9
• The instructor demonstrated a thorough knowledge of the course material
  • 5: 9/9
• fostered a good learning environment
  • 5: 9/9
• was well prepared and organized
  • 5: 9/9
• What are the best aspects of this course?
  • going through test case
  • very well written applicable software
  • a truly great instructor
  • the instructor
  • philosophical orientation toward modeling
  • material on nonlinearity and interaction; case studies; computing facilities
• Suggestions for future:
  • correct functions that didn't work (aregImpute, summary.areg.boot)
  • add survival analysis strategies
  • slow down, more time on lab examples and libraries
  • more work w/Design
  • little more info on package/software itself
  • faster (and deeper) on statistical theory

June 2-4 2003, Right Source, Philadelphia

1=disagree 5=agree

• The course met the objectives advertised on Insightful's websight and flyers
  • 5: 6/7 4: 1/7
• has provided me with a good foundation for working with S-PLUS
  • 5: 4/7 4: 1/7 3: 1/7
• The instructor demonstrated a thorough knowledge of the course material
  • 5: 6/7 4: 1/7
• fostered a good learning environment
  • 5: 6/7 4: 1/7
• was well prepared and organized
  • 5: 6/7 4: 1/7
• What are the best aspects of this course?
  • Hearing the instructor's assessment and advice about sources of information as well as about the intracacies of the statistical methods covered. It is also very good that datasets, scripts and other information are posted on Dr Harrell's website(s)
  • Seminar-like interaction with the instructor. Real-life examples and advice using state-of-the-art methods
  • Dr Harrell's teaching style and enthusiasm made it easy to ask questions and truly learn the material
  • Dr Harrell is one of the best teachers - internationally!
  • Frank Harrell is an excellent teacher! His breadth and depth of experience analyzing complex data are unsurpassed; he's practical and down to earth; admits to "logical inconsistencies" and is open to simple and complex questions
  • New methods for validation; diagnostic plots, encouraging me to try something new
• Suggestions for future:
  • Offer Part II - continued modeling strategies
  • Would like to have had more discussion of survival data; multivariate responses

ASA Continuing Education Program 6 Aug 2005, JSM, 1-day course

Summary of evaluations (frequencies)

Aspect	A	B	C	D	F
Presentation quality	28.5	8.5	3	0	1
Overall assessment	25	12	3	0	1

Comments

• Would have liked more examples
• A bit less time at beginning on generalities
• Very informative - appreciate the instructor's willingness to keep the course interactive
• Would rather receive the book and other material ahead of time
• Very beneficial - wish course was two days
• Less emphasis on model notation to allow more time on advanced topics
• Presentation lacked clarity and consistency and logical sequence. Wish topics were addressed in more depth with more examples
• Covered too many topics with controversial explanations. Not too convincing about rationale behind the proposed theory. Most important chapters with case studies were covered with a lot of hand waving.
• This is a nicely conducted course
• While not pleased to learn that much of what I've been taught is wrong, I enjoyed the perspectives presented and look forward to reading the book
• More examples; instructor knew his stuff very well
• Nice ideas to try in future but time and money might limit putting into practice, especially in an environment where one works on many projects at once
• Course was mostly an academic exercise and examples seemed to be of little practical value. Models are overly complex.
• Very in depth, knowledgeable and thoroughly covered
• Excellent material; a lot covered
• Excellent presentation! The instructor showed impressive mastery of the material. I was most impressed with his ability to answer "left field" questions authoritatively without hesitation, while conveying that he understood exactly what was asked!
• Inconsistent regarding use of descriptives; talks about not using before modeling, then gives case studies that start with descriptives
• Examples (real life applications) help to make the theory a little more understandable. Clear concise notes were nice.
• Excellent presentation, clear, fascinating and enjoyable
• Might break up the lecture with parts of case studies to better reinforce material rather than waiting until the end
• The instructor did an excellent job. His knowledge of the subject matter, humor, and presentation style made the course enjoyable and informative.
• I prefer the lecture format for these
• The instructor was plenty knowledgeable regarding the topic but I have little experience using these techniques. So these topics are irrelevent to my work.
• I am extremely skeptical of great manipulations of data to fit models (e.g., transformations, splines, imputation), probably because I work generally with small data sets. More consideration of science/statistics in the face of small data sets would be extremely beneficial. I also work in a Bayesian framework for most analyses, and understanding the principles of this course in a Bayesian framework would be useful.
• This was a great course! Frank was an excellent teacher. I'd recommend this course to all levels of statisticians. This course was very practical. I will be able to apply the course materials to my job immediately. * Very good. Thanks! I would like an example that could be used to illustrate the points throughout the class.
• Very good course

ASA Traveling Course May 2006

• Mid-Michigan Chapter, Lansing, 11May06

ASA Continuing Education Program 6 Aug 2006, JSM, 1-day course

Summary of evaluations (frequencies)

Aspect	A	B	C	D	F
Presentation quality	26	15	1	0	0
Overall assessment	21	20	1	0	0

Comments

• Excellent - very practical and impressively knowledgeable
• Suggest more examples in interpreting results
• Very good and enlightening course
• Great course instructor
• Need less breadth and more depth; very knowledgeable; good presentation skills
• Provocative instructor; poor handouts for a short course
• Use Powerpoint instead with more examples, and separate the topics a bit more
• Need more real-world examples
• An excellent course, but a little too broad in the material covered. I would want a bit more depth on fewer topics.
• It would be nice to have a couple of your sets to work with (?)
• Excellent and very informative
• It was a very useful course that should be offered again
• Excellent instructor
• I found this course very helpful. I hope that Dr Harrell is able to communicate some of his perspectives to a younger audience to help bring up the next generation properly!
• A lot of material for one day. Great instructor - good presentation and willing to discuss differing viewpoints. Very patient with class. Too much specific to R and not SAS or SPSS. They are industry-specific.
• Almost perfect lecture! Highly valuable; material could be better structured (sometimes difficult to discern what's important); specialized language often used; would be helpful to have more definitions
• A bit fast for me - I need a semester! and a bit much. Good presentation style and interacts well with audience.
• The handout sometimes are a little confusion (sic). Format a little better. Make some statements more clear.
• Very nice, examples very useful
• Good analytic advice. Goes overboard criticizing software packages.
• Fantastic topics and very practical
• The last 2 examples are appreciated
• EXCELLENT!
• Excellent presentations and material / topics selected
• Examples restricted to Medical. This course topics (sic) needs to be taught by someone with background in other areas as well [Instructor note: the Titanic and 1992 elections are medical?]

XLSolutions Two-Day Course with R/S-Plus Workshop 28-29Sep06

Comments on What are the Best Aspects of the Course

• Comprehensive strategy for modeling
• Good material - well paced, instructor stopped for questions and discussion - encouraged "student" participation
• Strategies, evaluation of methods, using cutting edge or most correct techniques
• The instructor was very good. Course content was relevant and well-taught. It was good that the focus remained on the methods vs. actual syntax
• Very knowledgeable instructor who is a very good speaker (esp. conveying important ideas); showing of R/S+ code for variety of issues, with computers for hands-on
• That Frank Harrell was teaching it

Comments on What Aspects Should be Changed

• Explain concepts more
• None obvious
• It's hard to keep up with the instructor's work in R - a little more explanation of syntax would be useful
• The course should give out Harrell's book to attendees!
• A bit more time for exercise could be accomplished with slightly more class time per day; length of days a bit short; software should be pre-loaded next time; possibly include non-health care examples as available and some with smaller sample sizes

Cleveland ASA Chapter 21Mar07

• Evaluations

ASA Continuing Educating Program, Joint Statistical Meetings, Denver, 2Aug08, 1-day course

Summary of evaluations (frequencies)

Aspect	A	B	C	D	F
Presentation quality	29	11	3	0	0
Overall assessment	22	17	4	0	0

Comments

• Like most short courses, the material was presented very fast; screen did not display the entire page
• It would have been better if he had mixed examples with the section to make more sense
• Time is not enough to cover all the material
• Time distribution to the content
• Course material was useful, but poorly presented
• Speaker is not able to control Q&A. A lot of material is skipped. Speaker makes biased comments on some statistical softwares (sic)
• The stuff covered is too much, maybe can be reduced some and emphasize some more practical and important things next time
• Try to give more practical examples w/hands on computer software use
• Too much emphasis on predictive modeling, neglecting discussion of effect estimation models; instructor extremely knowledgeable, good instructor
• The course was very good as was the instructor - very easy to understand. My only criticism is that I thought we spent too much time on splines, so it seemed that we had to rush through section 4. Although that is probably due to the many questions that were asked about splines. But overall a very good course.
• Should be a two-day course - too much wonderful things (sic) need to be skipped due to the shortage of time
• Manage audience questions; extremely useful
• The coverage can be chosed (sic) more carefully. The morning session was excellent but too many jumps in the afternoon when trying to cover more stuff.
• Too much negative commentary on stepwise techniques. It's fine that he doesn't like the technique (for a number of legitimate reasons) but he brings it up too often.
• More R code to do needed procedures
• I needed to read the book, which I am buying. I have been out of the field too long. Dr Harrell knows his field.
• Instructor tried to answer too many questions from the audience which makes it slow the presentation. Otherwise a good course.
• This course was an eye-opener for me! I will definitely attempt to learn more about regression modeling strategies.
• Excellent! Wonderfully useful!
• We did not cover all the notes - which was fine, we can refer to them later; except that notes are terse - they need some explanation. Dependent on software packages.
• Should be 2-3 days. Better graphical way to explain RCS. Best course EVER attended. Answers many questions why models and studies don't validate well. Gives me tools to create robust models. Dr Harrell talks at a very high level, perhaps loses some attendees, like myself, on a few subjects. The interaction section was too fast to grasp.
• Excellent instructor; very good course
• Dr Harrell is knowledgeable, confident, explains concepts well. Goes quickly, but this is necessitated by time limits. I appreciate the way he handled a few people asking too many questions and taking up too much time!
• I would like to see more case studies in finance & consumer behavior instead of concentration in biostat. Modeling/sampling related to large samples (more than 10 million obs).
• Enjoyed the class. I think a schedule of breaks and pauses should be presented at the beginning of the class. Material covered was fascinating and relevant to my work. I think the 2 case studies were excellent and maybe need to be covered earlier in the class.
• Many of us have SAS only for software, how work around?; consider starting case studies earlier to keep us interested.
• Excellent course. I would gladly take this course spread out over two days so there would be more time for examples.
• Good info to take away - practical rules of thumb w/explanations/reasons why; very understandable level - appreciate all the examples + R code; As w/"most" CE courses, started out at a slower pace & then had to pick up spped to cover material.
• The instructor is extremely knowledgeable and answers questions well. The course seems best suited to biomedical studies. I think the course attempted to cover too much material and therefore didn't cover anything very well or in depth. The course was helpful on a ceptual level, but even the case studies were covered so quickly that the main intent or ideas weren't reinforced or illustrated.
• Would be helpful if he worked some results out during the session to support some statements
• Great speaker
• Although the speaker has earned the right to proclaim his opinion, it was hard to distinguish between 'soapbox' and 'statistical validity/efficiency'

ENAR 2009 1-day Short Course, San Antonio TX

• Click here

ASA Continuing Educating Program, Biometric Section, Joint Statistical Meetings, Vancouver, 1Aug10, 1-day course

Summary of evaluations (frequencies)

Aspect	A	B	C	D	F
Presentation quality	24	13	1	0	0
Overall assessment	18	18	2	0	0

Comments

Most of the comments that were given were suggestions to spend a bit less time on introductory material and more time on case studies. Several participants thought too much material was covered for one day.