(c)2006 Albert Donnay

u__Results from
larger studies__ with more subjects
are more significant

than results from smaller studies (all other factors being equal).

*Beware of small
studies claiming great significance !
*

than results from smaller studies (all other factors being equal).

u__Studies with
control groups __are better than ones
without !

uThe __P-value__ specifies the
probability of obtaining a result as or more or extreme than the result observed due to chance alone,

*so small P-values
(less than or <.05=5%) are good, but P<.01 is better !
*

uThe __95% confidence interval__ (CI) specifies a range for the (unknown)

true result that is compatible with the (known) observed result.

It means that*if a study could be repeated 100 times, *

*the true result
would fall within the confidence interval 95% of the time.
*

true result that is compatible with the (known) observed result.

It means that

uAn __Odds Ratio__ (OR) is the odds of
a having particular outcome or disease given a particular exposure or other factor (such as a genetic
abnormality) compared to the odds
of having the same outcome without the exposure.

*An OR>1.5 is
usually considered significant, but >2 is very significant *

*(it means the
odds in the exposed are TWICE those of the unexposed)*