I am enclosing a histogram where variable age has multiple modes. Is there a specific way to report it.
Thanks!
You seem to have somehow introduced a measurement error in a part of your survey responses: some of people reported their exact age in years, and others reported the (nearest?) multiple of 5 (i.e. 30, 35, 40, etc.) Once you deal with this problem, the problem of multiple modes will go away.
What you see is very likely an artifact in the way the age was asked/given/recored. Obviousely there is some subset of people who have their age with a precision of 5 years (20, 25, 30, 35, ...), whereas others gave the age with a presicion of 1 year.
I am not sure how to best solve this "problem". My idea was to round all the age values to their nearest multiple of 5 and use and present this data.
Off-topic: Funnny, I honestly have not seen Alex' post at the time I wrote my answer that seems to be a plagiate of Alex' answer. It is not (I would not have taken the effort to repeat what is already said...).
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude