4 ms·

You are confusing statistical significance with practical significance. A difference of 0.00000001% can be statistically significant with enough data, but it wi

You are confusing statistical significance with practical significance. A difference of 0.00000001% can be statistically significant with enough data, but it will never be practically significant.

mensetmanusman 5y ago
Except in physics where the charge of an electron couldn’t be different by even less than that or else there could be no life. https://en.m.wikipedia.org/wiki/Fine-structure_constant https://en.m.wikipedia.org/wiki/Fine-structure_constant
Blikkentrekker 5y ago
And what exactly would be the meaning of the word “practical significance” then? what exactly would it signify? Because honestly; I feel this secondary meaning is a buzzword transferred from the first meaning which actually has meaning, but in the secondary case it's simply used to use an expensive word without understanding it.
- Jensson 5y ago
  Significance was a word before statistics used it to define one of their concepts. Just because a field takes a word people used in everyday contexts and used it to define something doesn't mean that the everyday meaning disappears. Edit: Another example is significant digits. Significant digits is a much older concept than significance in statistics, and maps much better to the original meaning of significance. Basically significant things are things you can't ignore when you do your analysis.
  - Blikkentrekker 5y ago
    > Significance was a word before statistics used it to define one of their concepts. Just because a field takes a word people used in everyday contexts and used it to define something doesn't mean that the everyday meaning disappears. That doesn't answer what I asked: what exactly is the meaning here because I remain unconvinced that it's not simply an empty buzzword inserted that transferred from it's usage of statistics, especially when used in conjunction with percentages. > Edit: Another example is significant digits. Significant digits is a much older concept than significance in statistics, and maps much better to the original meaning of significance. Basically significant things are things you can't ignore when you do your analysis. Indeed, and that has a meaning that I understand which is also morphologically transparent, but in this supposed meaning of the word which you've to elaborate upon it is neither morphologically transparent, nor do I actually see any meaning in it and I still feel it's used as a buzzword to simply make use of a fancy word for it's own sake.
    - benchaney 5y ago
      What on earth are you talking about? Significant is plain English word with a straightforward meaning. How is it "fancy" or a "buzzword"? As for what it means: "important and deserving of attention; of consequence" (https://www.dictionary.com/browse/significant https://www.dictionary.com/browse/significant)

You are confusing statistical significance with practical significance. A difference of 0.00000001% can be statistically significant with enough data, but it will never be practically significant.

mensetmanusman 5y ago

Except in physics where the charge of an electron couldn’t be different by even less than that or else there could be no life. https://en.m.wikipedia.org/wiki/Fine-structure_constant https://en.m.wikipedia.org/wiki/Fine-structure_constant

Blikkentrekker 5y ago

And what exactly would be the meaning of the word “practical significance” then? what exactly would it signify? Because honestly; I feel this secondary meaning is a buzzword transferred from the first meaning which actually has meaning, but in the secondary case it's simply used to use an expensive word without understanding it.

Jensson 5y ago
Significance was a word before statistics used it to define one of their concepts. Just because a field takes a word people used in everyday contexts and used it to define something doesn't mean that the everyday meaning disappears. Edit: Another example is significant digits. Significant digits is a much older concept than significance in statistics, and maps much better to the original meaning of significance. Basically significant things are things you can't ignore when you do your analysis.
- Blikkentrekker 5y ago
  > Significance was a word before statistics used it to define one of their concepts. Just because a field takes a word people used in everyday contexts and used it to define something doesn't mean that the everyday meaning disappears. That doesn't answer what I asked: what exactly is the meaning here because I remain unconvinced that it's not simply an empty buzzword inserted that transferred from it's usage of statistics, especially when used in conjunction with percentages. > Edit: Another example is significant digits. Significant digits is a much older concept than significance in statistics, and maps much better to the original meaning of significance. Basically significant things are things you can't ignore when you do your analysis. Indeed, and that has a meaning that I understand which is also morphologically transparent, but in this supposed meaning of the word which you've to elaborate upon it is neither morphologically transparent, nor do I actually see any meaning in it and I still feel it's used as a buzzword to simply make use of a fancy word for it's own sake.
  - benchaney 5y ago
    What on earth are you talking about? Significant is plain English word with a straightforward meaning. How is it "fancy" or a "buzzword"? As for what it means: "important and deserving of attention; of consequence" (https://www.dictionary.com/browse/significant https://www.dictionary.com/browse/significant)