Re: Is this enough information to make an inference?

On Fri, 28 Dec 2007 13:09:39 -0800 (PST), "michalchik@xxxxxxx"
<michalchik@xxxxxxx> wrote:

A link has been discovered between a certain virus (adeno-virus 36)
and obesity. Casual mechanisms are being elucidated biochemically but
I was wondering if there are any general statemenst we can make about
the virus's contribution to the obesity epidemic.

Given:

1) The virus is a causal risk factor for obesity and that obesity is
not a risk factor for catching the virus.

2) 20% of the general population tests positive for the virus.

3) 40% of obese people test positive for the virus.

4) 30% of the united states is obese.

Can we infer?

1) What percantage of obesity in the general population is
attributable to the effect of the virus.

The first step to inferring is to construct a table
of relative frequencies for a population of 100%.
You have the marginal proportions for both Virus and Obese,
and you have a cell, so you can get the other cells
by subtraction.

You can make crude estimates from that. - How may
"extra" cases are there in the Virus/Obese group?
Google to see if there is a better formula.

"Attributable risk" would probably be underestimated,
all things being equal, because there is surely a lag
between getting the virus and developing obesity.


2) What is the likelyhood that you will become obese if you contract
the virus.

I'd probably look at the Odds Ratio of the table
as an indicator of effect size. I'd want to see tables
for various ages... along with "duration of infection."

--
Rich Ulrich
http://www.pitt.edu/~wpilib/index.html
.

Relevant Pages