Karen Rosenbloom writes:
>I would like to ask the lists opinion about whether
>a 5 point Likert scale should be considered ordinal
>or nominal. I have found support of both camps.
I presume that you mean ordinal versus interval.
From the perspective of someone who is reviewing existing literature, the
answer is easy. There is support for both camps, so you should presume that
either analysis is credible.
From the perspective of someone who is analyzing their own data, the answer
is trickier. You might want to adopt the philosophy that is most likely to
be held by the referees of the journal you plan to submit your research to.
Another approach is to analyze the data as if it were ordinal and then
re-analyze it as if it were interval. If the two analysis provide the same
general conclusions, great! If they disagree, then you have just opened up
an avenue of further investigation for your data.
Steve Simon, [log in to unmask], Standard Disclaimer.
STATS: STeve's Attempt to Teach Statistics. http://www.cmh.edu/stats
Watch for a change in servers. On or around June 2001, this page will
move to http://www.childrens-mercy.org/stats
P.S. My personal opinion is to place Kilert scale items on an interval. Life
is short and an interval analysis is usually much easier to digest. That is
more important to me than rigidly insisting on an assumption of ordinal
scaling.
|