Statistics class today : Am I correct or is my teacher?

[oraclewizard77]

I got an A in statistics. So not to be too harsh on you, but on a bell shaped curve you would be on the bottom of your class.

So lets look at the example, out of 50, only 29 crystals were handed out. So its not even, and therefore there is bias.

If all 50 crystals got handed out, it would not matter if one person got more white crystals than blue crystals.

 

[Algo_Design_Kid]

Quote from oraclewizard77:

I got an A in statistics. So not to be too harsh on you, but on a bell shaped curve you would be on the bottom of your class.

So lets look at the example, out of 50, only 29 crystals were handed out. So its not even, and therefore there is bias.

If all 50 crystals got handed out, it would not matter if one person got more white crystals than blue crystals.

Dude, then you didn't get an A in reading comprehension. That wasn't the question. The question was that someone stated there was "bias" because some people got more than others.

My only point is that it doesn't make a difference who got them because their fractions would all add up to one in the end.

 

[oraclewizard77]

Total handed out = 29 out of 50

P1 3/9 b 6/9 w
P2 0/10 b 10/10 w
p3 2/5 b 3/5 w
p4 1/2 b 1/2 w
p5 3/3 b 0/3 w

b = 9/29 w = 20/29

No where do we see it add up to 1.

Even if you are simplistic you could add 9/29 to 20/29 to get 1, but we note that the original amount was 50 and 29/50 does not add up to 1.

 

[Algo_Design_Kid]

Quote from oraclewizard77:

Total handed out = 29 out of 50

P1 3/9 b 6/9 w
P2 0/10 b 10/10 w
p3 2/5 b 3/5 w
p4 1/2 b 1/2 w
p5 3/3 b 0/3 w

b = 9/29 w = 20/29

No where do we see it add up to 1.

Even if you are simplistic you could add 9/29 to 20/29 to get 1, but we note that the original amount was 50 and 29/50 does not add up to 1.

are you serious? maybe you are not understanding - we already know all the crystals arent being taken out. that is not the question

for example you could take someone with 6 blue crystals and split them up into 2x ( 3/6 ) to two other people. there would still be six crystals ( 3/6 + 3/6 )

does that make sense? if you go back to the OP the only argument is someone says there is SAMPLE SIZE BIAS and that was affecting the results. my only point this whole time is you can split them up all the ways you want but you will still sum the same amount of total crystals. so there can be no such thing as sample size bias in this example is my only statement

 

[oraclewizard77]

No you are wrong. You can not make up your own shit in your head. Statistics is hard science. It does matter that that 29 out of 50 crystals were taken out of the box, and that more white crystals were handed out than blue crystals. If all 50 crystals were handed out, it would not matter assuming true random if one person got more white crystals than blue crystals because the fractions would still add up to 1. Since this did NOT occur, we have sample bias.

Quote from Algo_Design_Kid:

are you serious? maybe you are not understanding - we already know all the crystals arent being taken out. that is not the question

for example you could take someone with 6 blue crystals and split them up into 2x ( 3/6 ) to two other people. there would still be six crystals ( 3/6 + 3/6 )

does that make sense? if you go back to the OP the only argument is someone says there is SAMPLE SIZE BIAS and that was affecting the results. my only point this whole time is you can split them up all the ways you want but you will still sum the same amount of total crystals. so there can be no such thing as sample size bias in this example is my only statement

 

[Algo_Design_Kid]

those were just trivial amounts in the OP. dont focus on those.. should have made that more apparent

 

[Algo_Design_Kid]

Quote from oraclewizard77:

No you are wrong. You can not make up your own shit in your head. Statistics is hard science. It does matter that that 29 out of 50 crystals were taken out of the box, and that more white crystals were handed out than blue crystals. If all 50 crystals were handed out, it would not matter assuming true random if one person got more white crystals than blue crystals because the fractions would still add up to 1. Since this did NOT occur, we have sample bias.

No, I am not wrong. You are wrong ( I think ).

Check this out.

If I randomly take crystals ( this wasn't a perfect random sample in our class ) and I just start handing them out it doesn't matter who get's how many of any crystals.

Let me try to explain again.

I randomly select from a box of balls and hand these out.

Person 1 - 3 white balls
Person 2 - 3 blue balls
Person 3 - 3 white balls

Would you agree that there is 6 total white balls?
Would you agree that there is 3 total blue balls ( lol ) ?

Lets say that I decided to only hand person 3 two balls this time and i will decide to hand 1 ball to person four.

Person 1 - 3 white balls
Person 2 - 3 blue balls
Person 3 - 2 white balls
Person 4 - 1 white ball

Would you agree that there is 6 total white balls ( like the previous example? )
Would you agree that there is 3 total blue balls ( like the previous example? )

Does it matter who gets how many balls as long as the sample remains exactly the same? This is my point, it doesn't matter to whom the balls got distributed to.

Please agree.





All that matters is that they will always sum up to the same amount of crystals.

 

[zedDoubleNaught]

Oh man, I'm so sorry to stir up the heated debate ... maybe there's another way to approach this ... the teacher said you would find out in Ch 5, what's the title or some of the subtitles from that section? Maybe one of them will be the name, and I can look it up on wikipedia.

My guess with "sample size bias" is there is bias that cannot be avoided any time you take less than the total population, but that's a memory from a course a long time ago, so I could have my concepts wrong. I thought it was fun to try to answer, and wanted to find out what the teacher and text book said was the answer, to see if I got it right.

 

[Algo_Design_Kid]

Quote from zedDoubleNaught:

Oh man, I'm so sorry to stir up the heated debate ... maybe there's another way to approach this ... the teacher said you would find out in Ch 5, what's the title or some of the subtitles from that section? Maybe one of them will be the name, and I can look it up on wikipedia.

My guess with "sample size bias" is there is bias that cannot be avoided any time you take less than the total population, but that's a memory from a course a long time ago, so I could have my concepts wrong. I thought it was fun to try to answer, and wanted to find out what the teacher and text book said was the answer, to see if I got it right.

OK, but she said it was in reference to people HAVING MORE THAN OTHER PEOPLE. I can see that definition, but I haven't looked it up personally. If that really is the definition I would agree.

But it wasn't in the context on the question / classroom discussion.

 

[oraclewizard77]

Exactly which is why as I pointed out, it is so crucial that 29 out of 50 crystals were handed out instead of 50 out of 50, but I guess it takes a book to prove who is right to the other poster instead of plain common sense.

Quote from zedDoubleNaught:

Oh man, I'm so sorry to stir up the heated debate ... maybe there's another way to approach this ... the teacher said you would find out in Ch 5, what's the title or some of the subtitles from that section? Maybe one of them will be the name, and I can look it up on wikipedia.

My guess with "sample size bias" is there is bias that cannot be avoided any time you take less than the total population, but that's a memory from a course a long time ago, so I could have my concepts wrong. I thought it was fun to try to answer, and wanted to find out what the teacher and text book said was the answer, to see if I got it right.