# statistics question

WatchPage 1 of 1

Go to first unread

Skip to page:

How do I solve part c? I thought of it along the lines of inverse binomial distribution where-

X~B(120, 5/120)

P(X=x)=0.95

How do I find x?

Any help would be appreciated

X~B(120, 5/120)

P(X=x)=0.95

How do I find x?

Any help would be appreciated

Last edited by mmike2003; 3 weeks ago

0

reply

Report

#2

so firstly the sign should not be equal but smaller or equal, and then you use the cumulative thing in your calc and you should find it that way by looking for the first one which is above 0.95, so in this case 9 i think

1

reply

(Original post by

so firstly the sign should not be equal but smaller or equal, and then you use the cumulative thing in your calc and you should find it that way by looking for the first one which is above 0.95, so in this case 9 i think

**dmears55**)so firstly the sign should not be equal but smaller or equal, and then you use the cumulative thing in your calc and you should find it that way by looking for the first one which is above 0.95, so in this case 9 i think

1

reply

Report

#4

**dmears55**)

so firstly the sign should not be equal but smaller or equal, and then you use the cumulative thing in your calc and you should find it that way by looking for the first one which is above 0.95, so in this case 9 i think

mmike2003

You have an open ended situation here. It is NOT a binomial distribution.

0

reply

(Original post by

Agree about the sign, but not with the rest of it.

mmike2003

You have an open ended situation here. It is NOT a binomial distribution.

**ghostwalker**)Agree about the sign, but not with the rest of it.

mmike2003

You have an open ended situation here. It is NOT a binomial distribution.

0

reply

Report

#6

(Original post by

Oh wait, I thought I use binomial distribution. Any hint? Or could you show some working for help please? Thanks

**mmike2003**)Oh wait, I thought I use binomial distribution. Any hint? Or could you show some working for help please? Thanks

If we let p be the probability that he wins in a given week, then:

What's the probability that he doesn't win in the first week?

Ditto in the first two weeks?

And the first three?

See the pattern. Then consider n weeks and form an inequality.

Last edited by ghostwalker; 3 weeks ago

0

reply

(Original post by

It's easiest to look at the complementary event. So, we're looking for the probability of him not winning in n weeks, say, being less than 0.05

If we let p be the probability that he wins in a given week, then:

What's the probability that he doesn't win in the first week?

Ditto in the first two weeks?

And the first three?

See the pattern. Then consider n weeks and form an inequality.

**ghostwalker**)It's easiest to look at the complementary event. So, we're looking for the probability of him not winning in n weeks, say, being less than 0.05

If we let p be the probability that he wins in a given week, then:

What's the probability that he doesn't win in the first week?

Ditto in the first two weeks?

And the first three?

See the pattern. Then consider n weeks and form an inequality.

The probability that he wins on any given week is 5/120

So, the probability that he doesn't win on any given week is 1- [5/120] = 23/24

how do I find week two and three? Once I get these I think I can find the pattern to solve the inequality

0

reply

Report

#8

(Original post by

Thanks for the hint.

The probability that he wins on any given week is 5/120

So, the probability that he doesn't win on any given week is 1- [5/120] = 23/24

how do I find week two and three? Once I get these I think I can find the pattern to solve the inequality

**mmike2003**)Thanks for the hint.

The probability that he wins on any given week is 5/120

So, the probability that he doesn't win on any given week is 1- [5/120] = 23/24

how do I find week two and three? Once I get these I think I can find the pattern to solve the inequality

Are weeks 2 and 3 any different to week 1, in terms of the probability of not winning?

Note: We're not just interested in just week2, but rather the first two weeks together, etc.

0

reply

(Original post by

Don't know the previous answers, so I assume your figures are correct.

Are weeks 2 and 3 any different to week 1, in terms of the probability of not winning?

Note: We're not just interested in just week2, but rather the first two weeks together, etc.

**ghostwalker**)Don't know the previous answers, so I assume your figures are correct.

Are weeks 2 and 3 any different to week 1, in terms of the probability of not winning?

Note: We're not just interested in just week2, but rather the first two weeks together, etc.

so week two would be (23/24)^2=0.918??

and week three would be (23/24)^3=0.8801?

Am I thinking along the right lines?

0

reply

Report

#10

(Original post by

Yes, I would assume it is given by (23/24)^n , where n is the number of weeks???

so week two would be (23/24)^2=0.918??

and week three would be (23/24)^3=0.8801?

Am I thinking along the right lines?

**mmike2003**)Yes, I would assume it is given by (23/24)^n , where n is the number of weeks???

so week two would be (23/24)^2=0.918??

and week three would be (23/24)^3=0.8801?

Am I thinking along the right lines?

Etc.

And hence the probablity of failing in all the first n weeks (i.e. not winning) is (23/24)^n, as you have.

So, it just remains to form the inequality and solve.

0

reply

(Original post by

The probabilty of failing in week2 is still 23/24, so the probablity of not winning in the first two weeks is (23/24)^2

Etc.

And hence the probablity of failing in all the first n weeks (i.e. not winning) is (23/24)^n, as you have.

So, it just remains to form the inequality and solve.

**ghostwalker**)The probabilty of failing in week2 is still 23/24, so the probablity of not winning in the first two weeks is (23/24)^2

Etc.

And hence the probablity of failing in all the first n weeks (i.e. not winning) is (23/24)^n, as you have.

So, it just remains to form the inequality and solve.

(23/24)^n<0.05

so, n=log base 23/24 (0.05)= 70.3890 . This rounds to 71 weeks I guess

Is this correct?

0

reply

Report

#12

(Original post by

I need to find the number of weeks to achieve a probability of winning at 95%, hence a probability of not winning at 5%

(23/24)^n<0.05

so, n=log base 23/24 (0.05)= 70.3890 . This rounds to 71 weeks I guess

Is this correct?

**mmike2003**)I need to find the number of weeks to achieve a probability of winning at 95%, hence a probability of not winning at 5%

(23/24)^n<0.05

so, n=log base 23/24 (0.05)= 70.3890 . This rounds to 71 weeks I guess

Is this correct?

And you should be aware that the distribution you're looking at is the Geometric.

1

reply

(Original post by

Looks good.

And you should be aware that the distribution you're looking at is the Geometric.

**ghostwalker**)Looks good.

And you should be aware that the distribution you're looking at is the Geometric.

0

reply

X

Page 1 of 1

Go to first unread

Skip to page:

### Quick Reply

Back

to top

to top