Many of you are still probably confused after our discussions of p values and confidence intervals yesterday, so I am enclosing a pamphlet which explains this pretty simply. The simplest possible way to describe the two is that the p value tells you the likelihood of an experimental result occurring just by chance ( as in the coin flipping in class, when we got three or four heads consecutively). A p of less than .05, shows that there is a less than 5 percent probability that a particular result occurred by chance. This p value is considered a minimum level for studies.
Confidence intervals tell you the range within which the true value can be found . For example, a result such as 31 ( 25-40) 95% CI, means that the study came up with the figure 31, the true value is between 25 and 40. Why is there such an range of values? Because each experiment or set of observations is from a particular sample ( ie 6,500 people in the case of the pravastatin case we discussed yesterday), and samples can vary. The CI tells us if we keep doing the same experiment with different samples, 95% of the time we will get a value between 25 and 40.
Still confused? Then just remember to look for P<0.5 in the results of a paper, and also look at the range for the 95% CI so you know the range within which the true value lies.
Don’t forget to finish the HIV/AIDS online course before next class, so we can have a good discussion.
Thanks for a good discussion on penicillin, antibiotic resistance etc yesterday and see you on Tuesday.