Names of places in Croatia

[FlatAssembler]

I don't understand what you mean. By p-value, I mean the probability of that k-r pattern in Croatian river names occurring by chance. By measuring collision entropy of the Croatian language and using the Monte Carlo birthday calculations, I have shown that the probability of that pattern occurring by chance is somewhere between 1/300 and 1/17.

But you’re continually touting (or at least implying) that p-values support your study hypothesis. P-values are only meant to test the statistical likelihood of a null hypothesis. But even if that was your null, it’s pretty pointless, since phonic patterns never arise by chance. If it did, the English word for ‘fish’ could be something like ‘grzblk’.

Boru

The mainstream linguistics claims that the river names that start with k-r (Krka, Korana, Krapina, Kravarščica, Krbavica and Karašica) are unrelated. That is the null-hypothesis. Which has been made unlikely by my calculations.

EDIT: Can you explain the part of your post that I bolded? I don't understand it?

[The Grand Nudger]

Right, so that is what you're asking, just in the context of a continued assertion or argument.  The short answer is no, and obviously no.

How is it "obvious"? I don't see it. If that were true, p-values would not be valued in social sciences (or natural sciences, for that matter), right?

I'm confident that you either can or will be able to understand why other people not making the same mistake you make doesn't make them more wrong than you.  Honestly, the question doesn't even make sense does it?

[FlatAssembler]

How is it "obvious"? I don't see it. If that were true, p-values would not be valued in social sciences (or natural sciences, for that matter), right?

I'm confident that you either can or will be able to understand why other people not making the same mistake you make doesn't make them more wrong than you.  Honestly, the question doesn't even make sense does it?

What is that "mistake"? An attempt to calculate the p-values?

[The Grand Nudger]

I feel like that's a question you should ask this guy?

even if it is wrong, isn't it still less wrong than the vast majority of papers in the field of Croatian names of places, since the vast majority of papers in that field do not even attempt to calculate the p-value?

isn't a paper that doesn't even attempt to calculate a p-value way more alarming in that regard than a paper that has a weak p-value?

[FlatAssembler]

I feel like that's a question you should ask this guy?

even if it is wrong, isn't it still less wrong than the vast majority of papers in the field of Croatian names of places, since the vast majority of papers in that field do not even attempt to calculate the p-value?

isn't a paper that doesn't even attempt to calculate a p-value way more alarming in that regard than a paper that has a weak p-value?

I don't see the point you are making. Are you against calculating p-values in the study of Croatian toponyms? Do you think there can be a science without statistics?
If you think there can be a science without statistics, look at the history of medicine. Medicine was basically worse than useless before the advent of statistics in medicine. Physicians were wrong on just about everything, and were probably killing more people than they saved.

[The Grand Nudger]

You apparently don't see the point you were making, either, eh?

[FlatAssembler]

@BrianSoddingBoru4 If you think I am not taking phonotactics into account, I am. I tried, in that paper, to estimate how much phonotactics takes away from the collision entropy of the consonant pairs in the Croatian language. It is slightly more than 1.5 bits per consonant pair.

EDIT: Yeah, I remembered it almost correctly. It is 1.623 bits per consonant pair. I don't know about the rest of the phonology, which is where the uncertainty in the p-value (that it is between 1/300 and 1/17) comes from.

[FlatAssembler]

You apparently don't see the point you were making, either, eh?

My point is that p-values are science. A paper which calculates p-values, like my paper, is scientific, even if it is wrong. A paper that doesn't calculate p-values is not scientific.

[arewethereyet]

[The Grand Nudger]

You apparently don't see the point you were making, either, eh?

My point is that p-values are science. A paper which calculates p-values, like my paper, is scientific, even if it is wrong. A paper that doesn't calculate p-values is not scientific.

Fractal wrongness appears to be the name of the game at this point.