[podcast]http://traffic.libsyn.com/llvlcshow/llvlc47-ned-kock.mp3[/podcast]
file size: 31.5mb
Ned Kock, author of the “Health Correlator” blog, is our guest today on The Livin’ La Vida Low-Carb Show with Jimmy Moore!
In addition to being a popular blogger, Ned Kock is a professor of IT at Texas A&M. Ned has a lot to say, and boy do he and Jimmy cram a lot of great content into today’s show! Find out more about Ned at his “Health Correlator” blog and snag a free download of his uber-geeky but oh-so-cool HealthCorrelator™ spreadsheet, a handy dandy tool he designed for people to analyze scientific data using his knowledge of both IT and livin’ la vida low-carb! ENJOY this very special episode.
ALSO: Special thanks to our friends at Quest Protein Bars! Get your TWO FREE BARS today!
LINKS MENTIONED IN EPISODE 477
– Support our sponsor: Low-Carb Quest Protein Bars
– Ned Kock bio
– “Health Correlator” blog
– NedKock.com
– HealthCorrelator™ for Excel software
And as always…
Add the show to your iGoogle or Google Reader page
Add the show in iTunes (No ipod required!)
Disclosure: http://cmp.ly/3
Cool Jimmy! I now see the little noise at the beginning that you told me about.
I wrote a post on my blog linking to this interview, and ended up getting more comments there than I expected:
http://bit.ly/kBgKMT
Hopefully more folks will come comment here.
Ned
Cool! You were awesome buddy.
Jiiiimmmmmy! This is the second interview in a row that made my head hurt! Man are they smart! I will have to listen to them a few more times to glean all the points. But for now I’m thinkin’ if LC is what they are doing, COUNT ME IN! Great interviews! I’m so happy for you!
THANKS! My head hurts from them sometimes too. 😀
This was my favorite interview this year so far.
Awesome Peter! He certainly packed a lot of info in an hour.
Another outstanding guest, Jimmy. Thanks for bringing them to us.
I’m glad Ned followed up on tumor necrosis factor later in the interview. I’ve forgotten what he said though, so now I’ll have to re-listen. ;-}
He said so much, my head was spinning…in a good way.
This really was an information packed interview with Ned. The comments on his blog are educational as well.
Thank you for listening Kathy!
Ned,
I work in industry and use Design of Experiments to create a design space for multiple factors at multiple levels, then do a partial least squares reqression for multiple responses to determine the relationship between the factors and responses. For example, with nutrition one factor could be carbohydrate types at two levels like refined sugar vs vegetables. Another factor could be carbohydrate amounts at two or three levels like 20 grams/day, 100 grams/day and 300 grams/day. You could create a very large design space by adding more factors like other macronutrients, types and levels of exercise and other critical factors like sleep. Responses would be things like blood sugar, blood pressure, blood lipid tests, body weight, percent body fat, etc. I’m not sure how the experiments would be executed, like with people or rats? I can see how it would become quite expensive as a large amount of data would need to be collected for a reasonable analysis. The regression coefficients would show the relationships between factors and responses. Factors with strong effects on responses could be determined from the coefficients and interactions between factors could also be determined. For example, hypothetically maybe high levels of dietary fat may have modest effects on percent body fat and aerobic exercise may have modest effects on percent body fat, but the combination of the two could have a strong effect on percent body fat.
I’m about half way through GCBC and haven’t seen much discussion on the design, math or analysis of these nutritional experiments. I’m interested in reading more about this type of discussion and seeing more rigorous math being applied to nutrition. What are your thoughts on this?
Kelly
Hi Kelly. It seems like what you have in mind can have quite an impact on our understanding of health issues. The area of health-related data collection and analysis, broadly speaking, is still lagging behind other areas in terms of use of techniques like PLS regression, resampling, and estimation of nonlinear effects.
I’d recommend that you define your predictors, as much as possible, as numeric variables, with a reasonable degree of expected variation (as opposed to categorical variables). This way you’ll be able to explore nonlinear relationships using software like WarpPLS.
Thank you all for your kind words!
Excellent! I consistently read Ned’s blog and it was a treat hearing him speak.
Great interview! I’ve known Ned as an information systems researcher for many years. I interviewed him for his first academic position and was impressed with his breadth and depth of knowledge in information systems. He brings this same careful, thoughtful analysis to issues of health and provides the evidence to back up his positions.
Hey John! Nice to see you commenting here, and thanks for the kind words.
My best to you and your family, Mr. Cool. I’ll call you soon …
Best, Ned
Hey, Jimmy – I can’t see any of the comments on any of the podcasts. 🙁 Am I doing something wrong?