Question 1157959: I am learning about the probability of independent events. I just need to know why do we multiply the probabilities. I want to be able to explain the reason if I am asked.
Found 3 solutions by ikleyn, Edwin McCravy, jim_thompson5910: Answer by ikleyn(52778) (Show Source):
You can put this solution on YOUR website! .
Two events, A and B, are called independent if P(A and B) = P(A) * P(B).
Here (A and B) is the event (A and B appear simultaneously).
So, two events are independent, BY THE DEFINITION, if and only if P(A and B) = P(A) * P(B).
----------------
To learn more on this subject, look into the lesson
- Independent and mutually exclusive events
in this site.
Thank you for asking (!)
Come again to this forum soon to learn something new (!)
Answer by Edwin McCravy(20054) (Show Source):
You can put this solution on YOUR website!
There are exceptions to the rules
"AND means multiply probabilities"
and
"OR means add probabilities".
What you asked for is WHY this always works:
P(A and B) = P(A)∙P(B [when we know that A is the case]) =
P(A and B) = P(B)∙P(A [when we know that B is the case])
If B is the case a third of time, and A is the case half of the time (that is,
when we know that B is the case), then half of those times when B is the case, A
will also be the case. So they both are the case simultaneously half of a third
of the time, so we multiply their probabilities and get (1/3)(1/2) = 1/6, so
they both will be the case AT THE SAME TIME 1/6th of the time.
Caution: When A rules B out, we may NOT multiply their individual probabilities,
for the probability that they are BOTH the case is ZERO, because they cannot
both happen at the same time.
Now we need to know the three kinds of pairs of events that are talked about in
probability studies. We must learn and understand and be able to distinguish
them:
1. A pair of (mutually) INDEPENDENT events
2. A pair of (mutually) DEPENDENT events
3. A pair of MUTUALLY EXCLUSIVE events.
[The word "mutually" is seldom used in the first two cases but is always used
in the third case.]
If a pair of events are such that one event DOES NOT INCREASE or DECREASE the
probability of the other event, then the events are INDEPENDENT.
If a pair of events are such that one event INCREASES or DECREASES the
probability of the other event, then the events are DEPENDENT.
A special case of a pair of DEPENDENT events is the case when one event
DECREASES the probability of the other event all the way to ZERO. In other words
one event COMPLETELY RULES THE OTHER EVENT OUT! Then the pair of events are
MUTUALLY EXCLUSIVE.
[For this last case, think of the less common use of "EXCLUSIVE" as 'EXCLUDING'.
One event EXCLUDES the other]
-------------------
The two formulas that ALWAYS work in ALL three cases are
(1) P(A or B) = P(A) + P(B) - P(A and B)
(2) P(A and B) = P(A)∙P(B|A) = P(B)∙P(A|B), where the "|" means "given".
In the case of a pair of INDEPENDENT events, formula (2) above can be
simplified.
In the case of a pair of MUTUALLY EXCLUSIVE events, formula (1) above can be
simplified:
In cases of pairs of INDEPENDENT events A, B,
P(A given B) = P(A|B) = P(A)
and
P(B given A) = P(B|A) = P(B)
so (2) above simplifies to
(2a) P(A and B) = P(A)∙P(B) in that case.
In the case of pairs of MUTUALLY EXCLUSIVE events,
P(A and B) = 0, so (1) simplifies in that case to
(1a) P(A or B) = P(A) + P(B)
Caution: Don't get any of the three cases mixed up! And, I repeat, remember
that there are exceptions to the rules "AND means multiply probabilities" and
"OR means add probabilities".
Edwin
Answer by jim_thompson5910(35256) (Show Source):
You can put this solution on YOUR website!
This is a great question. Consider we have a 4x4 grid as shown below
X | A1 | A2 | A3 |
A4 | A5 | A6 | A7 |
A8 | A9 | A10 | A11 |
A12 | A13 | A14 | A15 |
There are 16 items total. One of which is X. The other 15 are A1 through A15. The probability of randomly selecting X is 1/16
Now consider this grid
We have 9 items. One is Y, the other 8 are B1 through B8. The probability of randomly selecting Y is 1/9
I claim that the probability of selecting both X and Y at the same time, again completely randomly, is 1/144 which is the result of multiplying 1/16 and 1/9. Both events are independent as neither grid interacts with one another.
This is not obvious why this is the case. But let's say we arranged the items in the first grid to be along the left side of a very large table (16 rows) and the items of the second grid to be along the top of the large table (9 columns). We have this blank table
| Y | B1 | B2 | B3 | B4 | B5 | B6 | B7 | B8 |
X | | | | | | | | | |
A1 | | | | | | | | | |
A2 | | | | | | | | | |
A3 | | | | | | | | | |
A4 | | | | | | | | | |
A5 | | | | | | | | | |
A6 | | | | | | | | | |
A7 | | | | | | | | | |
A8 | | | | | | | | | |
A9 | | | | | | | | | |
A10 | | | | | | | | | |
A11 | | | | | | | | | |
A12 | | | | | | | | | |
A13 | | | | | | | | | |
A14 | | | | | | | | | |
A15 | | | | | | | | | |
I'll fill out part of the table. The first row shows
XY, XB1, XB2, ... , XB8
| Y | B1 | B2 | B3 | B4 | B5 | B6 | B7 | B8 |
X | XY | XB1 | XB2 | XB3 | XB4 | XB5 | XB6 | XB7 | XB8 |
A1 | | | | | | | | | |
A2 | | | | | | | | | |
A3 | | | | | | | | | |
A4 | | | | | | | | | |
A5 | | | | | | | | | |
A6 | | | | | | | | | |
A7 | | | | | | | | | |
A8 | | | | | | | | | |
A9 | | | | | | | | | |
A10 | | | | | | | | | |
A11 | | | | | | | | | |
A12 | | | | | | | | | |
A13 | | | | | | | | | |
A14 | | | | | | | | | |
A15 | | | | | | | | | |
Notation:
XY = we selected X and Y
XB1 = we selected X and B1
XB2 = we selected X and B2
etc etc
The rest of the table is filled out the same way. You just use the headers as shown. Anyways, note how there are 16 rows and 9 columns giving 16*9 = 144 items total in this massive table. Ask yourself: how many items are XY? That would be one only. This shows the probability of getting XY randomly is 1/144.
------------------------------------------------
As you can see, this can be thought of as an area problem. Areas of rectangles are found by multiplying the length and width. Squares are a special type of rectangle. Consider a square of area 1. This represents 100% probability.
Draw a square with area 1. Mark the upper half as a blue rectangle. If we throw a dart, then the probability of landing in the blue shaded region is 1/2. This is because 1/2 of the area is blue, and the total area is 1, so (1/2)/1 = 1/2.

Square divided into 2 equal pieces
Draw another square of area 1. Divide the rectangle into 3 vertical equal slices (think of a candy bar being broken up perhaps).

Square divided into 3 equal pieces. The probability of landing in the red region is 1/3.
If we overlap the two figures, then we'll have the blue and red regions combine to form the purple area

Square divided into 6 equal pieces
Ask yourself this: what is the probability of landing in BOTH the upper blue area and the red area at the same time? That would be 1/6. The area of the blue rectangle is (1/2)*1 = 1/2. The area of the red rectangle is 1*(1/3) = 1/3. So the purple region has area (1/2)*(1/3) = 1/6
Or you can think of it like this: Start with a red rectangle (area 1/3) and cut it in half to get the purple rectangle (area 1/6). A slight variation: Start with a blue rectangle (area 1/2) and cut that into thirds to get the purple piece (area 1/6)
|
|
|