Talk:Guttman scale
![]() | Statistics Unassessed | |||||||||
|
it seems that the sample guttman scale is of little use, as the first 4 categories are all such that each is implied by the subsequent (i.e. allowing immigrants next door automatically implies you'd allow them in your neighborhood [unless you're in a weird border case where nextdoor isn't in your neighborhood]).
- That's the whole point of a Guttman scale, and that is the required structure of the data. Holon 07:15, 29 June 2006 (UTC)
What a horrible mess!!!
This article now begins like this:
-
- A hypothetical, perfect Guttman scale consists of a unidimensional set of items that are ranked in order of difficulty from least extreme to most extreme position. For example, a person scoring a "7" on a ten item Guttman scale, will agree with items 1-7 and disagree with items 8,9,10. An important property of Guttman's model is that a person's entire set of responses to all items can be predicted from their cumulative score because the model is deterministic.
What in hell does that mean? "difficulty" seems to mean there "items" are some sort of tasks that someone is to do. That's just a guess. A guess! Why should the reader have to guess? Then "a person scoring a "7" on a ten item Guttman scale".... So a Guttman scale is something on which someone scores points. Why can't the article just say so instead of leaving us to surmise this. Is this about sports? Tennis, maybe? Or chess? Then: "An important property of Guttman's model is that a person's entire set of responses to all items can be predicted from their cumulative score because the model is deterministic." Really? So someone named "Guttman" has a "model". I still don't know if this is about tennis or weighlifting or academic performance on physics examinations or what.
What a mess. Michael Hardy (talk) 17:15, 22 April 2008 (UTC)
Actually I disagree with your view. While the article is not 100% in readability or especially good at explaining in layman's terms all the aspects of a Guttman scale it does give the reader with some minimal statistical insight a very good understanding of a Guttman scale and even manages the explain more advanced concepts such as coefficients and their relation to the test performed.
For the record, a Guttman scale can be applied to tennis, weightlifting, academic performance and so on (although you might have to create some weird statistics to require it). It's not a scale for ranking specific performances but rather "sets of items in order of difficulty" (as the article says). I suggest going back and reading the introduction again before reading the article...