Return to Website

Number Watch Web Forum

This forum is about wrong numbers in science, politics and the media. It respects good science and good English.

Number Watch Web Forum
Start a New Topic 
Author
Comment
Non-Risk Ratio

This has to be a calculation mandated by the epidemiologists, right. I have never seen it though.

Here's how I would calculate it.

Our Trojan number is 10,000.

5,000 people who do an activity (ingesting orange juice for example.)

5,000 people who do not do that activity.

we get 30 people performing the activity who come down with the ailment (toenail cancer).

we get 20 people not performing the activity who come down with the same malady.

If I understand my Epidemiology properly (unlikely), this gives me a Risk ratio of 1.5 (30/20).

The non risk ratio 4970 / 5000 for people performing the activity or 99.4% of not getting the malady.

for those not performing the activity it would be 4980/5000 or 99.6% of not getting the malady.

4970/4980 give the non-risk associated with performing the activity or 99.8%.

The actual increased risk of performing the activity is 1-(non-risk) or 0.2%.

I suspect that even this number would often be dubious.

For some reason, I have always found the risk ratio disturbing. It reminds me of the presentation Statistics of the question what is probability of not rolling a 6 if you roll 6 dice. Perhaps I am unique in my first pass at this question, but I tried using 1/6 to get to the answer. Everyone here is likely aware that the answer is really 1-(5/6)^6. My impression of RR is that it is 1/6 answer to a 1-(5/6)^n question.

Is my reasoning suspect?

Re: Non-Risk Ratio

My apologies ahead of time if it turns out I read this in Sorry, Wrong Number and am now regurgitating it as if it were my own. I know that the roots of it are definitely in those pages as well as the pages of Junkscience Judo. Both of those books are at home right now, and in this fast paced age of Soccer and Baseball (my child is involved in both, making my wife's free time minimal), the chances of me getting home, fixing dinner, playing and getting kids ready for bed will likely drain me of any desire to negatively confirm the non-existence of such a thought within those page.

The above is an example of how rational people lose. Qualifying every word we speak to make sure we aren't taken out of context, then having the qualifying words used against us.

oh well.