What is correct terminology when describing the number of response levels of a Likert scale? I know that the Likert scale is the sum (or some sort of combination) of all the responses of a survey, test etc.  and the actual numeric responses are called response levels.  Is this true of all psychometric numeric responses?  ie are those responses also called response levels or are they some times called a scale (appropriately called that is)?
If you have any literature to back up what you say it would be appreciated.
 A: I'm not sure if I'm in a position to write a dictionary definition for these terms, but perhaps the following will help. And there probably are other terms used. But here's a basic overview:


*

*Scale: The measurement tool that results from taking a function of component items.

*Scale Score: The actual score of an individual on the scale.

*Item or question: An individual item that forms part of the scale.

*Rating scale or response scale: The set of response options available. (e.g., the set {"Yes", "No", "Don't Know"})(note also that this is sometimes, somewhat confusingly, also referred to as a scale)

*Response option or level of the response scale: An element  (e.g., "Strongly Disagree") of the response scale.

*Observation or response: the actual response (i.e., one of the response options) given by a particular participant to a particular item.


And here's it all used in a paragraph:

John gave the response "strongly agree" to item 7 on the Happiness scale.
  "Strongly agree" was the response option chosen. This was one of five options that made up the rating scale for each item. 
  Taking the mean of all the items on the scale yielded a scale score of 3.7.

