Is the percentage of marks scored by someone or something like the population growth ratio scale of interval? Like both have an absolute zero so it makes sense for them to be ratio and not interval?
$\begingroup$
$\endgroup$
1
-
$\begingroup$ It is none of the above--and this is one of many reasons why it is rarely worthwhile considering what the measurement scale of a variable might be. A justification for my first remark is that in most applications a score is equivalent to its complement; e.g., getting 79 marks out of 100 is the same as missing 21 out of a hundred. Thus, if you were to maintain this scale had a zero level, then the complementary score would have a zero level, too: and that would be the same as a score of 100. None of Stevens' scale types has two distinct zero levels! Similarly, the scale cannot be of ratio type. $\endgroup$– whuber ♦May 8, 2022 at 13:09
Add a comment
|