Korean Job Discussion Forums

[tomato]

According to this Website, there have been 1200 dinosaur fossils uncovered.
According to M. L. Lubenow, author of Bones of Contention, there have been 4000 humanoid fossils uncovered.
If we are to believe what archeologists tell us, all of the dinosaurs were found in Mesozoic rock and all of the humanoids were found in Cenozoic rock.
This strongly implies that dinosaurs lived only in the Mesozoic times and humanoids did not appear until Cenozoic times.
However, this does not prove the case absolutely.
What, then, are the chances that there might be an uncovered Mesozoic caveman or Cenozoic dinosaur?

I am trying to make a chi-square contingency table according to this Website.

This:



is what I came up with.

I am to make the calculation:

5200 [(4000) (1200) - (0) (0)] squared / (4000) (1200) (0) (0)

No problem with numerator of 1.19808E+17, but we have a denominator of zero, and I know that can't be right.

The only solution I can think of, then, is to plant one dummy Mesozoic caveman and one dummy Cenozoic dinosaur:



This gives us the calculation:

5200 [(4000) (1200) - (1) (1)] squared / (4000) (1200) (1) (1)

Again, the Excel program gives 1.19808E+17 for the numerator, but 4800000 for the denominator.
Divide and you get 24,959,989,600.
Since there are two rows and two columns of data, there is 1 degree of freedom.
The alpha table only goes up to 10.827, which gives a probability level of .001.

My questions are:

--Am I using the right statistical test?
--Is there anything else I can do about the zero's in two of the categories?
--Is there an alpha table which goes high enough for this purpose?
--Is there any way to calculate the alpha level for this collection of data?

#########################################

Another possible way to analyze the data is to pretend that we are flipping a coin and that we correctly predicted 4000 heads and 1200 and made no incorrect predictions.

This calculation is simpler:

1. Add 4000 and 1200.
This gets us 5200.

2. Calculate 2 to the power of 5200.
My Excel program won't go that high.

Is there any way to calculate this?

[The evil penguin]

ARRRRRRGGGGGHHHHHHH!!!!!

My brain HURTS!!!!!!

[johndoe]

Can't you play computer games like the rest of us?

Or go ghost hunting?

http://forums.eslcafe.com/korea/viewtopic.php?t=96912

[midgic]

[Tony_Balony]

You don't need math help, you need inferential statistics help. If you said "inferential statistics help needed" I would have gone an looked at other much dumber posts.

[jeffkim1972]

I believe you're using the wrong model.

Before you apply the model, you must make some assumptions.

But in this case, are you trying to prove that there is a small chance that a dinasour lived in the Cenozoic period and vice versa? A statistical model cannot be used to disprove empirical data.

You must use binary variables in your model if you want to create one model to predict dinasour or human fossils (or remains).

x[0,1]= 1 ; if in the Mesozoic Period
x[0,1]= 0 ; if in the Cenozoic Period

z[x;(parameters for dinasour model), (parameters for human model)]=(model for dinasours)*x + (model for humans)*(1-x)

This is not exactly predicting another planet's presence using another's perturbed motion from gravity.

Since there is no theory, other than that a big meteorite killed the dinasours. Evidence being that increase in iridium, something like that.

Think of it this way. if you have red and green balls in one jar and black and white balls in another. No statistical model will give you a chance that a red ball can be found in the same jar as the black and white balls.