@Eagle Mum
This is gonna sound hella dumb but no dumb questions right?
If the P(AnB) = P(A) x P(B)
and P(A|B) = P(AnB)/P(B)
then substituting P(A|B) = P(A) x P(B)/P(B) which is just P(A) as it cancels out???
I tried to google the shaded regions of what the venn diagrams look like for P(AnB) and P(A|B) but got nowhere so any help would be great.
Also that stuff on Bayes theorem, is that syllabus or nah? TY
It’s always good to ask questions when you don’t understand something.
I note that Drongoski has already replied, but since you addressed me, my additional cent worth, is that a probability tree diagram is quite useful in questions like this.
For this particular question:
P(A) is the probability that the manager is present (0.39)
P(Ā) is the probability that the manager is absent (0.61)
P(B) is the probability of a defect
P(B̅) is the probability of no defect
Therefore, in the tree diagram:
In the first branch, P(A) = 0.39, P(B|A) = 0.02, P(A ∩ B) = 0.0078
In the second branch, P(A) = 0.39, P(B̅|A) = 0.98, P(A ∩ B̅) = 0.3822
In the third branch, P(Ā) = 0.61, P(B|Ā) = 0.03 P(Ā ∩ B) = 0.0183
In the fourth branch, P(Ā) = 0.61, P(B̅|Ā) = 0.97, P(Ā ∩ B̅) = 0.5917
Edited: note that the sum of all branches should add up exactly to 1.
The question states that the condition is there is a defect, so the denominator is the sum of the first plus third branches: (0.39 x 0.02 + 0.61 x 0.03), whilst the numerator is the first branch (the manger was present when the defect occurred) 0.39 x 0.02.
Therefore, the answer is 0.0078 / (0.0078 + 0.0183) = 26/87 = 0.2988...
WRT your question, the problem with using a Venn diagram, as Drongoski has replied, is that B is not independent of A (when the manager is present, the defect rate fell from 0.03 to 0.02), so in a Venn diagram, P(B) within the shaded area would be 0.02, whereas P(B) in the rest of the circle B outside of the shaded area would be 0.03. In theory, the four regions of a Venn diagram (the fourth being the area outside both circles) could represent the four branches of the probability tree, but the tree is much more suitable for visualising all the possible cases and the probability of each case.
A probability tree can have many more branches than the four shown here from two binary possibilities. To apply Bayes theorem, all you need to be able to do is pick the appropriate branch(es) for the denominator and for the numerator from the stated information (ie. conditions).