Selecting a Distribution for Orange Juice Taste Testing with N Judges

  • Context: Undergrad 
  • Thread starter Thread starter boneill3
  • Start date Start date
  • Tags Tags
    Distribution
Click For Summary

Discussion Overview

The discussion revolves around selecting an appropriate statistical distribution for a taste testing experiment involving N judges who evaluate four brands of orange juice. The focus is on determining the probability that at least one judge selects a specific brand (brand A) as the best choice, given a total of 10 judges. The conversation includes considerations for simulation of trials.

Discussion Character

  • Exploratory
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • Brendan poses a question about the best distribution to use for simulating the probability that at least one judge chooses brand A, given that the probability of any single judge choosing A is 1/4.
  • One participant suggests considering the probability that no judge chooses brand A as a starting point.
  • Brendan calculates the probability of not choosing A for one judge as 0.75 and expresses uncertainty about how to extend this to 10 judges, suggesting a binomial distribution might be appropriate.
  • Brendan attempts to apply the binomial distribution to calculate the probability of no judges picking A, leading to a very low probability of 0.000000953 for no judges selecting A.
  • Brendan later corrects his calculations and finds the probability of at least one judge picking A to be approximately 0.943, indicating a high likelihood.
  • Another participant confirms Brendan's revised probability calculation as correct.

Areas of Agreement / Disagreement

There is a general agreement on the use of the binomial distribution for this scenario, and Brendan's final probability calculation of 0.943 is confirmed as correct. However, earlier calculations and interpretations show some uncertainty and confusion, indicating a progression in understanding rather than a consensus from the outset.

Contextual Notes

Brendan's initial calculations included errors and misunderstandings about the application of the binomial distribution, which were later corrected. The discussion reflects the iterative process of refining mathematical reasoning in statistical contexts.

Who May Find This Useful

This discussion may be useful for individuals interested in statistical modeling, particularly in the context of experimental design and probability calculations in taste testing scenarios.

boneill3
Messages
126
Reaction score
0
I have been given a question that reads:
In a taste testing experiment N judges are offered four different brands of orange juice to taste.
Suppose Y is the number of of judges that choose brand A as the best.

I know that the probability of a judge choosing A is 1/4.

I have been asked what is the probability that at least one judge chooses brand A given that there are 10 judges.

My question is what would be the best distribution to use?

As I'm going to have to simulate 1000 trials.

Thanks
Brendan
 
Physics news on Phys.org
No need to simulate. Think of it this way... what is the probability that NO judge chooses A?
 
The probability of not A is 0.75 but that is for 1 judge, but the probability of 10 judges not picking juice A would be a lot lower.

I thought I might be able to use a binomial distribution to figure it out so I can simulate for any number of judges?
regards
Brendan
 
Just wondering if this makes sense.

Let x = 1 be the number of successes (here the success is that the judges don't pick A)
Let n = the number of judges = 10

So P (x) = 0.75

If I use the binomial distribution

P(x=0) = 10Choose0 * P(x)^x * P(not x)^n-x

P(x=0) = 10Choose1 * (3/4)^0 * (1/4)^10

P(x=0) = 0.000000953

In another words it would be extremely improbable to have no judge pick A
 
So would the probability of at least one picking juice A be P(1-P(x=0) = 0.000000953)

> 99.9

regards
Brendan
 
Sorry guy's I think I've got it wrong.
I'll try again!

Let x = 1 be the number of successes (here the success is that the judges pick A)
Let n = the number of judges = 10

So P (x) = 0.25

If I use the binomial distribution to find the probability of 10 judges not choosing A

P(x=0) = 10Choose0 * P(x)^x * P(not x)^n-x

P(x=0) = 10Chooseo * (1/4)^0 * (3/4)^10

P(x=0) = 0.056

So the probability of at least 1 must be (1 - 0.056) = 0.943

I hope this is right cause I'm loosing my mind!
Brendan
 
Yep, 0.943 is right.
 
Thanks mate!
 

Similar threads

  • · Replies 29 ·
Replies
29
Views
6K
Replies
1
Views
3K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 16 ·
Replies
16
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
Replies
5
Views
2K
  • · Replies 42 ·
2
Replies
42
Views
5K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 9 ·
Replies
9
Views
2K
  • · Replies 14 ·
Replies
14
Views
6K