Probability related to Normal Distribution

Click For Summary

Homework Help Overview

The discussion revolves around finding the probability that one normally distributed variable, B, is greater than another normally distributed variable, A. Variable A has a mean of 55 and a variance of 9, while variable B has a mean of 65 and a variance of 25.

Discussion Character

  • Exploratory, Assumption checking, Mathematical reasoning

Approaches and Questions Raised

  • Participants explore the meaning of the probability P(B > A) and question whether A and B are independent. Some suggest using integration to find the probability, while others express confusion about the integration process and how to start solving the problem.

Discussion Status

Some participants have provided guidance on considering the random variable Y = B - A and calculating its expected value and standard deviation. There is an ongoing exploration of how to approach the problem, with some participants indicating they are beginning to understand the concepts involved.

Contextual Notes

Participants note that previous questions they encountered involved specific numerical values, which adds to their confusion regarding this problem's setup.

songoku
Messages
2,512
Reaction score
393

Homework Statement


Variable A has mean 55 and variance 9, variable B has mean 65 and variance 25. If A and B are normally distributed, find P (B > A)


Homework Equations


z = (x - μ) / σ



The Attempt at a Solution


Can this be solved? What is the meaning of P (B > A)? The probability of random variable B is bigger than random variable A?

Thanks
 
Physics news on Phys.org
I suppose the problem assumes A and B are statistically independent. Then the formal way to do this is:
P(B>A) = \int_{b=-\infty}^{+\infty} \int_{a=-\infty}^{b}f_A(a)f_B(b)dadb
This is a very nasty integral.
 
songoku said:

Homework Statement


Variable A has mean 55 and variance 9, variable B has mean 65 and variance 25. If A and B are normally distributed, find P (B > A)


Homework Equations


z = (x - μ) / σ



The Attempt at a Solution


Can this be solved? What is the meaning of P (B > A)? The probability of random variable B is bigger than random variable A?

Thanks

Yes, P(B > A) means exactly what you said. It can be solved quite easily if A and B are independent, but I cannot give more details now: first you have to make a serious attempt to solve it yourself.
 
hikaru1221 said:
I suppose the problem assumes A and B are statistically independent. Then the formal way to do this is:
P(B>A) = \int_{b=-\infty}^{+\infty} \int_{a=-\infty}^{b}f_A(a)f_B(b)dadb
This is a very nasty integral.

Sorry I don't get the idea how to do the integration and why can it be like that

Ray Vickson said:
Yes, P(B > A) means exactly what you said. It can be solved quite easily if A and B are independent, but I cannot give more details now: first you have to make a serious attempt to solve it yourself.

I don't have idea how to start. The question I saw until now always contained number such as P (B > 23). Please give me clue how to start solving this problem. What is the first step I need to take? I can't find the value of z because I don't know the value of the random variable.

Thanks
 
songoku said:
Sorry I don't get the idea how to do the integration and why can it be like that



I don't have idea how to start. The question I saw until now always contained number such as P (B > 23). Please give me clue how to start solving this problem. What is the first step I need to take? I can't find the value of z because I don't know the value of the random variable.

Thanks

Look at the random variable Y = B - A.
 
Ray Vickson said:
Look at the random variable Y = B - A.

OK maybe I get it

Let: Y = B - A

E(Y) = E(B) - E(A) = 10
σ(Y) = σ(B) + σ(A) = 34

P(B > A) = P(B - A > 0) = P(Y > 0) = P(Z > -1.715) = 0.9568

Do I get it right?

Thanks
 
songoku said:
OK maybe I get it

Let: Y = B - A

E(Y) = E(B) - E(A) = 10
σ(Y) = σ(B) + σ(A) = 34

P(B > A) = P(B - A > 0) = P(Y > 0) = P(Z > -1.715) = 0.9568

Do I get it right?

Thanks

It looks right.
 
Ray Vickson said:
It looks right.

ok thanks
 

Similar threads

Replies
7
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 8 ·
Replies
8
Views
2K
Replies
1
Views
1K
  • · Replies 1 ·
Replies
1
Views
1K
Replies
2
Views
2K
Replies
2
Views
1K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 8 ·
Replies
8
Views
4K