Descrete case

[PHX Admin]

Example (C.) The data collected in a toxicology study are recorded in the table below

Treatment/Events	Yes	No
Placebo	0	40
Active	3	37

A test for association using Fisher’s exact test gives us a p-value of 0.24. The clinical scientist argues that the 3 events are clinically significant even though it is not statistically significant. He argues this based on his experience of not seeing events often.

To see how prior experience would change the results, suppose that the scientist bases his statements on a sequence of 100 subjects on placebo with no event. If such data is ‘appended’ to the current information the data becomes

Treatment/Events	Yes	No
Placebo	0	140
Active	3	37

with a significant p-value of 0.01.

How do you incorporate the relevant prior experience of the clinician in your analysis?

[PHX Admin]

I am not sure what to do with this question.

[PHX Admin]

Me again!

[Author]

Posts