Question

A research article reports the results of a drug test. This medicine used to reduce vision loss in people suffering from macular degeneration. The article state p = 0,04 section findings.
The following issues are of three different interpretations of this value p. Note which of these three explanations are correct and which are not.

a) Is the probability that the results are as or more extreme than those of the study, if the drug is not effective. (True or False)

b) Is the probability that the drug is not effective. (T/F)

c) Is the probability that the drug is effective.T/F

Answer 1

I believe that:
a) is False
b) is True
c) Is False.

What do you believe?

Answer 2

Thank you in advance!!

Answer 3

a is true, b and c are false. You can not get b or c with null hyothesis testing in frequentist statistics. See also: http://en.wikipedia.org/wiki/P-value

Answer 4

Thank you for your help Eugen.H!!

Answer 5

@Eugen.H But you din't tell me the explanation. In this group the explanation matters. Is not supposed to tell directly the answers. Be carefull with that please. We are here to learn not just to take answers!

Answer 6

@Eugen.H anyway thank you again you helped me a lot. I read many papers to learn the correct idea for p-value. For one thing I am sure. The most of researchers do not know the true meaning of p-value. I am very happy that I have now the best Professor in Statistics and I wish I had like him when I was an undergraduate. Thank you colleague!

Answer 7

I do think the big problem with null hypothesis testing (nht) is that there are two parts. One part is math and the second is decision theory and nobody talks about the decision theoretic part in earnest. In your example there is no decision theoretic part because the question is just the definition of the p-value and i do not have a better, or more elaborate answer than a.) or wikipedia. The big question is what to do with the p-value that you calculated. Typically you define an alpha value before and conclude that the null hypothesis is wrong if the p-value is less than alpha.

Answer 8

I agree with you. Nobody talks about the decision theoritic part. Do you believe that the researchers set up the alpha value before? I am afraid that the most researchers set up the alpha value after they see the p-value.

Answer 9

In social sciences the standard alpha is the magical 0.05 and natural sciences have higher standards so there is no possibility to arbitrarily set it to your p-value. There are similar ways to improve the chances of significance, tough.

Answer 10

0.05 is very common even and in natural sciences. Anyway thank you again for your information. Any extra knowledge is a gift. Jigoro Kano the founder of Kodokan Judo said "Nothing under the sun is greater than education. By educating one person and sending him into the society of his generation, we make a contribution extending a hundred generations to come."
All the best colleague!