If this is your first visit, be sure to
check out the FAQ by clicking the
link above. You may have to register
before you can post: click the register link above to proceed. To start viewing messages,
select the forum that you want to visit from the selection below.
Sorry I commented on your signature and not on your link my bad
"If man made it, don't eat it" - Jack Lallane
People say I am on a "crazy" diet. What is so crazy about eating veggies, fruits, seafood and organ meats? Just because I don't eat whole wheat and processed food doesn't make my diet "crazy". Maybe everyone else with a SAD are the "crazy" ones for putting that junk in their system.
P number: measure of how likely the difference seen is by chance or the study detected a real difference. OR odds ratio (similar to RR relative risk, but RR is measured in double blind, placebo controlled, OR is an estimate of RR for case control (this study), or cohort studies). So, for example, in this study:
a person who drank 1 bottle of wine daily has a 2.61x risk of having gastric cancer than someone who didn't. What are the chances that this risk occurred by chance? In other words, what is the likelihood that by chance this group of people were sorted in such a way that they seem to have this risk, which is a funtion of sorting and nothing else. That chance is given by the P number. Round 0.049 to 0.5 would equal a 5 out of 100 chance that this would happen by chance. Or, a 95 out of 100 that is wasn't by chance, i.e., a real effect. Generally, P<.05 is considered significant or real. The smaller the P, the more likely real.
Don't be fooled by P numbers alone, there is a lot more to consider when evaluating studies though. That's a long reply, so I'll leave it at that. For a good example, read Gary Taubes Good Calories Bad Calories.
cool, thanks for that. Obviously I'm not going to base a health choice on a single study (not that I'm about to start drinking an entire bottle of wine in an evening anyway). I was actually looking for a particular study that links grain/starch intake with gastric cancer (the Modan one) and came across this and was curious about what those cryptic numbers meant.
If we’re not supposed to eat animals, how come they’re made out of meat? Tom Snyder
the p value is not actually the probability of obtaining results by chance (that's the alpha value). The p value is the probability of obtaining these or more extreme results when the null hypothesis is true. I'm personally not a big fan of this statistic, as it really does not tell us that much, and is easily manipulated. I once analyzed a huge data set for a project, and got very low p values for an effect that was barely even considered small.
the odds ratio, in simple terms, is a description of association of probability between two groups. For instance, in the study you linked us to, uneducated people had 4.33 times the odds of developing cancer than educated people.
(I'm a social and quant psych grad student, it's my job to be a stats nerd).