GI Load, again, US population..

[Wuzzy]

This is my proof that pasta is a strong component, it is
almost as bad as ice cream (i don't know why i keep getting
bacon as being bad, obviously ppl who eat bacon eat alot of
high glycemic foods):

my p value is a perfect zero, beat that.

my correlation matrix of GI with foods: This proves that i
have done it right because as you can see The more you eat
coca cola, the more your glycemic score will be, v. coffee:
the more you eat coffee the score is still positive but not as
much as coca cola.

GI GI 1 cola 0.493656825 Cakes, cookies, brownies, etc
- times/mo 0.411952568 chocolate candy and fudge
0.405570424 potatoes 0.401332409 hi-c, tang, koolaid
etc 0.357834642 white bread/rolls 0.325336855 salted
snacks 0.322636068 Orange Juice 0.302494107 rice
0.295271855 Other Fruit Juices 0.261007305 corn bread,
muffins, tortillas 0.258370066 Citrus Fruits
0.254603232 Bacon/sausage/processed meats - times/mo
0.249233156 Ice cream,ice milk,milkshakes - times/mo
0.245673847 beef 0.234459381 beans/lentils/chickpeas
0.223943668 tomatoes 0.215641487 Cheese, all types -
times/month 0.210457249 Spaghetti/pasta w/ tomato
sauce-times/mo 0.205345929 Eggs 0.19874399 Other
Fruits 0.19754955 Pork+ham 0.19449729 Cheese dishes -
times/month 0.193948167 Stew or soup with vegetables -
times/mo 0.178799299 other veggies 0.175064673 Pizza,
calzone, lasagna - times/month 0.165559113
peanuts/peanut butter 0.162629848 hot cereals
0.153932355 Peaches 0.152782431 Carrots 0.136098313
sweet potatoes 0.132113555 Chicken/Turkey 0.125601795
hot red chili peppers 0.123407559 Cabbage, cole slaw,
sauerkraut-times/mo 0.120599915 Chocolate milk and hot
cocoa - times/mo 0.115805094 Liver and other organ
meats - times/mo 0.108849593 Spinach, greens, etc -
times/month 0.10385621 Melons 0.102716063 Milk to
drink or on cereal - times/month 0.097267353 other
cold cereals 0.091944701 Shrimp, clams etc 0.082148994
cereals: total, etc 0.079244935 dark bread/rolls
0.071121662 Brussels sprouts/cauliflower 0.057324285
flour tortillas 0.05721827 Fish 0.052528822 Yogurt and
frozen yogurt - times/month 0.028121972 coffee
0.010708107 cereals:all bran, etc -0.001602487 diet
cola -0.073244574

Now for a real manoever, I will determine what food
contributes the most to GI..(can anyone recommend a better
method?) I will do this using stepwise forward regression
maximum of 4 variables which will tell me the 4 most
significant predictors in decreasing order as follows:(if
anyone has a better method i would be glad to run it)

Y-Variable: GI

Method: Forward

Steps P R-Square Corrected

Ice crm 0 0.06847355 0.068364408 pasta 0 0.117727592
0.117520826 OrngJce 0 0.223640339 0.22336739 Bacon 0
0.259831209 0.259484201

Summary N R R-Square Std.Error normal 8537 0.509736411
0.259831209 12.03627526 corrected 0.509395918 0.259484201

Equation Coef Conf. (±) Std.Error T P Constan 24.07171895
0.452998012 0.231086462 104.64 0 Ice cream 0.40129309
0.035429728 0.018073657 22.202 0 Spaghetti/pasta 0.773597487
0.070202457 0.03581216 21.601 0 Bacon/processed 0.299006905
0.02869746 0.014639346 20.4244 0 Orange Juice 0.309872404
0.017453077 0.008903283 34.80 0

Analysis of variance SSq Degrees of Freed Mean Square F P
Regression 433905.959 4 108476.4897 748.7 0 Residue 1236047.24
8532 144.8719222 Total 1669953.199 8536 195.6365041

[Jay Tanzma]

Wuzzy wrote:
>
> This is my proof that pasta is a strong component, it is
> almost as bad as ice cream (i don't know why i keep getting
> bacon as being bad, obviously ppl who eat bacon eat alot of
> high glycemic foods):
>
> my p value is a perfect zero, beat that.
>
> my correlation matrix of GI with foods: This proves that i
> have done it right because as you can see The more you eat
> coca cola, the more your glycemic score will be, v. coffee:
> the more you eat coffee the score is still positive but not
> as much as coca cola.

> Equation Coef Conf. (±) Std.Error T P Constan 24.07171895
> 0.452998012 0.231086462 104.64 0 Ice cream 0.40129309
> 0.035429728 0.018073657 22.202 0 Spaghetti/pasta 0.773597487
> 0.070202457 0.03581216 21.601 0 Bacon/processed 0.299006905
> 0.02869746 0.014639346 20.4244 0 Orange Juice 0.309872404
> 0.017453077 0.008903283 34.80 0

Bacon doesn't make biological sense. It might be showing up
because its consumption is correlated with other high-GI
foods. I would rerun the analysis using stepwise selection
(you are apparently using forward selection -- there's no such
thing as "stepwise forward"). Bacon might get kicked out of
the model automatically. If it doesn't, then, one, check for
high univariate correlations between bacon and high GI foods,
and, two, force bacon out of the model, and see what,
variable(s) are selected instead, if any.

-Jay

[Jay Tanzma]

Thanks for formatting your tables better.