This is an example I use when I teach data types. It happens because the scale (F or C) is an “interval” scale. Its zero is not based on the absence of the property it is measuring, so you can’t apply a multiplicative transform to it like, “double”.
It is like lining up by height, calling the shortest person the standard and measure height of everyone else from that. So, the next tallest might be 2 cm, the next 4cm. But clearly the person we are calling 4cm is not twice the height of the person we called 2 cm.
Even if the scale was aligned with absolute zero - like Kelvin - it would not be able to describe temperature changes in the multiples primarily because our FEEL of temperature is what matters here. And since humans live in the approx. temperature ranges of -40 to 80 (using an extended range to cover cases like the Arctic/Antarctic stations, or saunas), the best scale to use would be a Celsius scale shifted somewhat to make 0deg the most optimal neutral temperature - which is, in my opinion, 16 degrees Celsius.
Ha, while funny it still doesn’t work. If we use an interval scale with zero degrees Lat defined as 16 degrees Celsius, how many times hotter is zero degrees Lat than-1 degrees Lat? If you are using “temperature comfort” as your underlying property, zero had to be the university defined “lack of all comfort” which I don’t think you will find. Subjective comfort is notoriously difficult to make into ratio scale. Pain measurement is a well- known example.
They live in Sweden so Celsius
It’s still an odd way to compare temperatures:
This is an example I use when I teach data types. It happens because the scale (F or C) is an “interval” scale. Its zero is not based on the absence of the property it is measuring, so you can’t apply a multiplicative transform to it like, “double”.
It is like lining up by height, calling the shortest person the standard and measure height of everyone else from that. So, the next tallest might be 2 cm, the next 4cm. But clearly the person we are calling 4cm is not twice the height of the person we called 2 cm.
Even if the scale was aligned with absolute zero - like Kelvin - it would not be able to describe temperature changes in the multiples primarily because our FEEL of temperature is what matters here. And since humans live in the approx. temperature ranges of -40 to 80 (using an extended range to cover cases like the Arctic/Antarctic stations, or saunas), the best scale to use would be a Celsius scale shifted somewhat to make 0deg the most optimal neutral temperature - which is, in my opinion, 16 degrees Celsius.
Ha, while funny it still doesn’t work. If we use an interval scale with zero degrees Lat defined as 16 degrees Celsius, how many times hotter is zero degrees Lat than-1 degrees Lat? If you are using “temperature comfort” as your underlying property, zero had to be the university defined “lack of all comfort” which I don’t think you will find. Subjective comfort is notoriously difficult to make into ratio scale. Pain measurement is a well- known example.
I mean thats completely valid criticism but it had a lot of shock value still.
Yes, the difference in temperature is interesting, I don’t want to seem dismissive of that. Just the choice of wording was also interesting.
yes
You’d think so, but without specifying the scale… it could be anything. ANYTHING!
Probably bananas though, according to my middleschool math teacher