Moivre-Laplace theorem (homework)

Click For Summary

Homework Help Overview

The discussion revolves around the application of the Moivre-Laplace theorem in the context of a problem involving Bernoulli trials. The original poster is tasked with using normal approximation to compute a probability related to the number of successes in 10 trials, each with a success probability of 0.7.

Discussion Character

  • Exploratory, Assumption checking, Problem interpretation

Approaches and Questions Raised

  • Participants discuss the interpretation of the absolute value equation and the correct formulation of inequalities. There is an exploration of the normal approximation method and its application to the problem, including the use of mean and standard deviation.

Discussion Status

Participants are actively engaging with the problem, questioning the validity of certain inequalities and discussing the implications of using a normal approximation for a discrete random variable. Some guidance has been offered regarding the proper setup of the inequalities and the potential need for a continuity correction.

Contextual Notes

There is mention of the challenges posed by the small sample size (n=10) and the limitations of the normal approximation in this context. Additionally, participants note discrepancies in the results obtained from different calculators, highlighting the complexity of the calculations involved.

Poetria
Messages
267
Reaction score
42
< Mentor Note -- thread moved to HH from the technical math forums, so no HH Template is shown >[/color]

I have no idea what I am doing wrong. Here is my problem:

S(n) - the number of successes in 10 Bernoulli trials, each with probability 0.7

I am supposed to use normal approximation to compute :

P{|S(n)/10-0.7|>0.2}

We are allowed to use a calculator.

I got:
|S(n)-7|>2|
5<S(n)>9

I have computed S*:

S(n)-7/1.44914
7=mean
1.44914- standard deviation

5*1.44914-7<Z<9*1.44914-7

Area (probability): 0.2451

But there is no such answer in the options.
 
Last edited by a moderator:
Physics news on Phys.org
Poetria said:
< Mentor Note -- thread moved to HH from the technical math forums, so no HH Template is shown >

I have no idea what I am doing wrong. Here is my problem:

S(n) - the number of successes in 10 Bernoulli trials, each with probability 0.7

I am supposed to use normal approximation to compute :

P{|S(n)/10-0.7|>0.2}

We are allowed to use a calculator.

I got:
|S(n)-7|>2|
5<S(n)>9

I have computed S*:

S(n)-7/1.44914
7=mean
1.44914- standard deviation

5*1.44914-7<Z<9*1.44914-7

Area (probability): 0.2451

But there is no such answer in the options.
What do you mean by 5<S(n)>9 ?

That's an invalid compound inequality.
 
  • Like
Likes   Reactions: Poetria
I see. :( My line of thought was as follows:

S(n)-7>2 or S(n)-7<-2 (this is an absolute value equation: |S(n)/10-0.7|>0.2)

How to correct it?
 
Poetria said:
I see. :( My line of thought was as follows:

S(n)-7>2 or S(n)-7<-2 (this is an absolute value equation: |S(n)/10-0.7|>0.2)

How to correct it?
What you have here is correct.

S(n)-7>2 or S(n)-7<-2

That says S(n) > 9 or S(n) < 5 . (You had S(n) > 5 )

You can't make these into a compound inequality.
 
Poetria said:
< Mentor Note -- thread moved to HH from the technical math forums, so no HH Template is shown >

I have no idea what I am doing wrong. Here is my problem:

S(n) - the number of successes in 10 Bernoulli trials, each with probability 0.7

I am supposed to use normal approximation to compute :

P{|S(n)/10-0.7|>0.2}

We are allowed to use a calculator.

I got:
|S(n)-7|>2|
5<S(n)>9

I have computed S*:

S(n)-7/1.44914
********************
Wrong: use parentheses, like this: (S(n) - 7)/1.44914

********************
7=mean
1.44914- standard deviation

5*1.44914-7<Z<9*1.44914-7

***********************
Wrong: you need 5 > 7 + 1.44914*Z or 7 + 1.44914*Z > 9

The complement of the event "{S(n) < 5} or {S(n) > 9}" is {5 <= S(n) <= 9}, which translates to
5 <= 7 + 1.44914*Z <= 9, or -1.38 <= Z <= 1.38 in the normal approximation.

However, since n = 10 is small, the normal approximation might not be very good, but it can be improved a lot by using the so-called "1/2-correction". The random variable S(n) is discrete (taking values 0,1,2,...,10 only), so {5 < S(n) < 9} = {S(n) = 6,7,8}. In the normal approximation, replace this by {5.5 <= S_normal <= 8.5}, so
5.5 <= 7 + 1.44914*Z <= 8.5. The resulting increase in accuracy is striking in this example.

*********************
Area (probability): 0.2451

But there is no such answer in the options.
Poetria said:
< Mentor Note -- thread moved to HH from the technical math forums, so no HH Template is shown >

I have no idea what I am doing wrong. Here is my problem:

S(n) - the number of successes in 10 Bernoulli trials, each with probability 0.7

I am supposed to use normal approximation to compute :

P{|S(n)/10-0.7|>0.2}

We are allowed to use a calculator.

I got:
|S(n)-7|>2|
5<S(n)>9

I have computed S*:

S(n)-7/1.44914
7=mean
1.44914- standard deviation

5*1.44914-7<Z<9*1.44914-7

Area (probability): 0.2451

But there is no such answer in the options.
 
Last edited:
  • Like
Likes   Reactions: Poetria
Thank you so much. I have tried to work it out. I think I got it. My problem is that I still can't get an exact answer. The closest (with the correction you have suggested):
if x = 5.5 I get 0.15031188 (I can choose 0.1675)
http://www.danielsoper.com/statcalc3/calc.aspx?id=53

The result: 0.08377348 is very imprecise.
Actually some calculators demand x, some Z. It is a mess.
 
Sorry, I get 0.71860226. Even worse.
 
I got it at last. :) Many thanks. :) You have helped me a lot. :) :) :)
 

Similar threads

Replies
10
Views
3K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 11 ·
Replies
11
Views
2K
Replies
15
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 7 ·
Replies
7
Views
4K
  • · Replies 22 ·
Replies
22
Views
4K
  • · Replies 5 ·
Replies
5
Views
5K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 9 ·
Replies
9
Views
4K