The very many questions-not-worth-their-own-thread question thread XXVII

[warpus]

Get her to eat some whipped cream, and wait until she has some on her ring finger. Then gently place her finger into your mouth and get the whipped cream off. Excuse yourself and run to your room, so that you can measure the hole in your mouth.

 

[Borachio]

Yeah, that'll work.

Spoiler :

Not.

 

[gay_Aleks]

Doesn't she, like, wear gloves?

 

[Borachio]

While eating whipped cream? That doesn't sound particularly lady-like.

Depends on the gloves though.

These?

Or these?

 

[gay_Aleks]

Well, obviously, the winter gloves might be a bit off, but it should give you a good idea of how big your future fiancée's finger is.

 

[The_J]

I think I have enough suggestions, thanks .
(didn't remember these Ring Pops though; does that stuff still exist? Would be perfect ^^)

And someone gets to make a new thread.
Who, Borachio or Warpus?
(I'll never get that right)

 

[Arakhor]

Which means...?

Since you asked...

 

[Borachio]

New thread.

 

[Sanguivorant]

Another math-based/comprehension question for you all:

"Matt thinks that he has a special relationship with the number 4. In particular, Matt thinks that he would roll a 4 with a fair 6-sided die more often than you'd expect by chance alone. Suppose p is the true proportion of the time Matt will roll a 4."

I am going to assume that the probability of rolling a 4 on a fair 6-sided die will be 1/6.

This is the proportion for rolling a 4.

However, Matt makes the claim that p > 1/6.

The alternative claim would be that p < or = to 1/6.

Is my reasoning correct?

I am mostly focused on my alternative claim, since that is what I will need in order to do my hypothesis test.

 

[Borachio]

I can't see how your reasoning isn't correct.

The probability of rolling a 4 with a fair 6-sided die is indeed 1/6. (By definition really. I don't see how Matt could think otherwise, tbh. If it's not so, then the die isn't fair.)

And if something isn't > than something else, it must be < or = to it. (Assuming that those comparators can be applied at all.) That covers the field, leaving no gaps.

 

[madviking]

Another math-based/comprehension question for you all:

"Matt thinks that he has a special relationship with the number 4. In particular, Matt thinks that he would roll a 4 with a fair 6-sided die more often than you'd expect by chance alone. Suppose p is the true proportion of the time Matt will roll a 4."

I am going to assume that the probability of rolling a 4 on a fair 6-sided die will be 1/6.

This is the proportion for rolling a 4.

However, Matt makes the claim that p > 1/6.

The alternative claim would be that p < or = to 1/6.

Is my reasoning correct?

I am mostly focused on my alternative claim, since that is what I will need in order to do my hypothesis test.

for hypothesis testing i think you would test both matt's hypothesis [p(4) > 1/6] and the null hypothesis [p(4) = 1/6].

otherwise, i'm not sure about your question

 

[Sanguivorant]

So the null hypothesis is thst p(4)=1/6? What makes you say that?

 

[madviking]

In statistical inference of observed data of a scientific experiment, the null hypothesis refers to a general statement or default position that there is no relationship between two measured phenomena

http://en.wikipedia.org/wiki/Null_hypothesis

in this case, the two phenomena are matt's affinity for fours and rolling a four at a higher rate

 

[Samson]

So the null hypothesis is thst p(4)=1/6? What makes you say that?

Yes. The point of a null hypothesis is that you can build a statistical model on it, compare the real results to that model and derive a probability of getting the results seen or more extreme given that the null hypothesis is true. If this probability is lower than some threashold (5% is frequently used) then you can reject the null hypothesis.

So in this case you say the null hypothesis is that p(4)=1/6. Given this we can build a model, and do an experiment (say 16 4's in 60 rolls), work out the probability of 16 or more 4's given p(4)=1/6 R: sum(dbinom(16:60, 60, (1/6))) = 0.0338462. As this is less than 5% we can reject the null hypothesis.

 

[dutchfire]

Moderator Action: Closed, new thread is here http://forums.civfanatics.com/showthread.php?t=540105