# How to decide on a cut-off score (groups) if the scale doesn't have one?

#### AbsolutelyNoIdea

##### New Member
For my study, I'm using a scale (rating scale from 1-7) that doesn't have a predetermined cut-off score. However, in order to do the stats, I need to make 2 groups/categories with "high score" and "low score". How do I determine the point at which a participant is considered high vs. low scoring? Can I just use the median or the mean, or 1 std. deviation below and above the mean as a cut-off point?
(Please no complicated mathematical procedures..Thanks)

#### hlsmith

##### Less is more. Stay pure. Stay poor.
Why do you need to dichotomize data?

#### noetsi

##### Fortran must die
Generally you don't dichotomize data when its not naturally dichotomous - you lose information. An obvious exception is dummy variables used to take pieces of a categorical variable [that does not seem what you are doing].

The answer to you question is you do this division on what makes theoretical/substantive sense. If you don't know this, I am not sure you can chose to divide the data this way.