PDA

View Full Version : SUMMARY: Likert scale



jdura50
05-21-2008, 05:36 PM
Enclosed you'll find the summary of the responses I collected to the
following question:

***ORIGINAL POSTING*****
Usually a Likert scales have 5 levels. Sometimes 7, 9 or 11.
But the final objective normally is reduced to 3 levels. For example:
1) People think that A is better than B
2) People think that A is similar to B
3) People think that A is worst than B
Recently I was discussing with a workmate if it is better to use 3
levels in our questionnaires.
My workmate thinks that 3 levels are more clear for elderly people and
time is saved.
Then, why researchers use 5 or more levels? Tradition?
I would like to find a scientific reference in order to justify that
the number of levels must be 5 or more.
******************************

The answer is that few levels reduce reliability. But many levels
increase fatigue. Then 5-7-9 levels are the usual.

References:
Carolyn C. Preston a, Andrew M. Colman (2000). Optimal number of
response categories in rating scales: reliability, validity,
discriminating power, and respondent preferences. Acta Psychologica,
104, 1-15.

http://www.amstat.org/sections/srms/Proceedings/papers/1986_091.pdf

http://epm.sagepub.com/cgi/content/citation/31/3/657

Many thanks to:
Eran Dayan
Melinda Whetstone




Saludos,
Juan V. DurĂ¡