# QM probability

1. Jun 30, 2006

### neutrino

For those who have the book, this is problem 1.4 from Griffiths, 2nd ed.

$$\psi (x,0) = \left\{ \begin{array}{rcl} A\frac{x}{a} & \mbox{for} & 0 \leq x \leq a \\ A\frac{b-x}{b-a} & \mbox{for} & a \leq x \leq b \\ 0 & otherwise \end{array}\right.$$

a) Normalise the wavefunction.
I found $A = \sqrt{\frac{3}{b}}$ (Am I right?)

c)
Where is the particle most likely to be found, at t =0?
Using the above value for A, I evaluated the two integrals, one from 0 to a, and the other from a to b. For the first, I get a probability of a/b, and for the second integral 1- (a/b), as expected. How does this answer the question?

2. Jun 30, 2006

### StatusX

Probably not. The answer should probably involve both a and b (or neither).

Where is the probability function maximized?

3. Jun 30, 2006

### HallsofIvy

Not even close! As StatusX pointed out, it should depend on both a and b. How did you get that?

It doesn't! Why would you think it would? Where is the wavefunction a maximum?

4. Jun 30, 2006

### neutrino

Ok here's what I did (I knew I hadn't understood how to do this properly :grumpy:)

$$\int_{0}^{a} (A\frac{x}{a})^2dx + \int_{a}^{b} (A\frac{b-x}{b-a})^2dx = 1$$ , for the wavefunction to the left of 0, and to the right of b being zero.
$$\frac{A^2}{3}(a + (b-a)) = 1$$

Last edited: Jun 30, 2006
5. Jun 30, 2006

### George Jones

Staff Emeritus
Before doing anything, I think sketching the wavefunction for this question is quite useful. Maybe this is already done in the book - I don't have my copy with me.

6. Jun 30, 2006

### neutrino

George, sketching is problem b.

7. Jun 30, 2006

### George Jones

Staff Emeritus
Looks OK to me.

8. Jun 30, 2006

### StatusX

I apologize, it does appear you were right, and A does not depend on a. The wavefunction superficially appears to be symmetric between a and b, but when you draw it you see they have very different roles, and the dependence on a cancels out when you integrate.

Last edited: Jun 30, 2006
9. Jun 30, 2006

### George Jones

Staff Emeritus
Here's roughly what's happening. The larger a is, the wider and shorter the wavefunction is, and this happens in such a way that area under the square of the wavefunction in independent of a. The width of the wavefunction also depends on b, but the height doesn't This leaves b in the final answer.

However, before doing the calculation, I would never have thought of this.

10. Jun 30, 2006

### neutrino

I feel a bit better now.

11. Jun 30, 2006

### StatusX

Clearly the integral of the wavefunction does not depend on a, because you are just taking half of the area of the rectangle of height 1 stretching from x=0 to x=b. More suprising is that the integral of the square of the wavefunction also does not depend on a, and this can be explained (and generalized) as follows.

Imagine integrating any monotonically increasing function from 0 to 1, and assume f(0)=0. This is equivalent to calculating the area under the curve y=f(x) on this region, and this in turn is equivalent to the area to the right of the curve up to the line x=1, which is just the integral from 0 to f(1) of 1-f-1(y). When you stretch the region of integration horizontally you are essentially mutiplying this latter integral (over y) by a constant, which explains why the value of the integral scales linearly with the amount of stretching. It isn't hard to show this extends to any function, not necessarily jut monotonically increasing ones.

In this problem, the amount of horizontal stretching is determined by a, and so in particular the integrals will be of the form k(a)+k(b-a)=kb, explaining why the dependence on a drops out. Sorry to dwell on this, but I find it interesting.

Last edited: Jun 30, 2006
12. Jun 30, 2006

### George Jones

Staff Emeritus
The probability that the particle will be found at any particular x between 0 and b is exactly zero. The question asks for for an x such that the particle is more likely to be found "in the neighbourhood" of this x than "in the neighbourhood" of any other x.

The probabilty density is given by the square of the wavefunction, so the particle is most likey to be found in the neighbourhood of the maximum of (the square of) the wavefunction.

13. Jul 3, 2006

### neutrino

Did you mean 1?

Thanks for the reply, George. (Sorry for being late with a reply)...I did plot $|\psi|^2$. It is discontinuous at the point $a$, but the graph also reaches a max at $\left(a, \frac{3}{b})\right$. So, is the answer to the question $a$ (around $a$)?

14. Jul 3, 2006

### nrqed

No, George *did* mean zero. Because |Psi(x)|^2 is a probability density so |Psi(x)|^2 dx represents the probability of finding the particle between x and x+dx. So strictly speaking, when people say at what x is the probability max, they really mean "for what value of x is it most likely to find the particle between x and x+dx?". But it is standard sloppiness to talk about "values of x for which the probability is max".
Well, the answer is simply "at x=a".

Patrick
EDIT: Again, the real question should have been "for what value of x is it most probable to find the particle between x and x+dx?". And the answer is "at x=a" (meaning that the probability is maximum for finding the particle between a and a+dx).

Last edited: Jul 3, 2006
15. Jul 4, 2006

### maverick280857

At t = 0, you already have the wavefunction. If you sketch it, you will see that it is a triangular section centered at x = 0 stretching from x = 0 to x = b and zero everywhere else. It is clear then that the wavefunction is maximum at x = a and so is the probability of finding it (probability "maximum").

16. Jul 4, 2006

### borisleprof

Your result for A is right.
You have to understand the meaning of Psi.
If, at the instant t, a measurement is made to locate the particle associated with the wave function Psi(x,t), the the probability P(x,t)dx that the particle will be found at a coordinate x and x+dx is equal to
Psi*(x,t)Psi(x,t)dx. Where Psi*(x,t) represent the complex conjugate of Psi(x,t) (Quantum physics, R. Eisberg, R. Resnick)

In the interval [0, a] Psi*Psi is increasing and in the interval [a,b] it's decrease to zero. So the maximum probability Psi*(x,t)Psi(x,t)dx is obtained around x = a;
So the particle is more likely to be found at x = a and x = a+dx.
bye

17. Aug 20, 2008

### Niles

I am dealing with this question too.

Why is it that the expectation value of x, <x>, is NOT equal to the point at which the particle is most likely to be found at t=0? And how is it that I can just use Psi, and not |Psi|2?

18. Aug 20, 2008

### Dick

If a particle has a 50% probability of being found at x=0 and a 50% probability of being found at x=1, then <x>=0.5. On the other hand, there's no chance it will actually be found at x=0.5. You can use Psi because the maximum of |Psi| is located in the same place as the maximum of |Psi|^2. Think about it.

19. Aug 21, 2008

### Niles

Does this mean that <x> is the point, where the probability of finding the particle to the right of <x> equals the probability of finding the particle to the left of <x>, which is 0.5?

About using |Psi|: Since |Psi| is the norm of Psi, then |Psi| and |Psi|^2 will have maximum at the same x - can this explanation be used?

Last edited: Aug 21, 2008
20. Aug 21, 2008

### HallsofIvy

No. it doesn't. <x> is the mean or expected value of the probability distribution. The value of x such that the probability is 1/2 of being less than x and 1/2 of being more than x is the median. They are not necessarily the saime.