When discussing a theoretical distribution having a single parameter described as a degree of freedom (like the $t$ distribution), do we refer to this parameter as degrees of freedom (plural) or degree of freedom (singular). We frequently see expressions such as A t test with 10 degrees of freedom. Could (should) we say A t test with degree of freedom of 10?

I feel that this parameter is a single concept, so discussing it should go with singular? As another example, consider The exact degree of freedom depends on ... Should I write here degree of freedom in its singular form? A reviewer commented that an "s" should be added.

Thanks for this linguistic aid.

  • 2
    $\begingroup$ It's a single parameter but it refers to, quite often, multiple degrees, so I would always use the plural. $\endgroup$ Sep 4 '20 at 18:27

It is plural, 10 degrees. That means 10 orthogonal dimensions in which the system may vary. This is much estabilished in many different fields. It's easy to find plenty of validation by just googling the term.

Also, the plural comes more natural when you act on the system to add or remove any number of degrees of freedom (fixing this parameter frees one degree of the model/adds one degree of freedom).

Addendum: dictionaries prefer the singular for the item title, but the usage should be plural, refer to Collins', where all the definitions start with "one of the..."


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.