Nearly 400 people participated in this study, and we had expected to find a single category within this sample. (This would make two distinct groups: category members, and category non-members.)
Instead, we found four distinct groups of people who deliberately hurt themselves.
Originally we tested our data with a statistical method that can detect the presence of a single category (that is, two groups). According to this method, either our data contained one category, or they did not.
When we did not find a single category, we tested our data with a different method that could tell us exactly how many groups we had. These groups are called clusters. They're like categories in that each group is distinct from every other group. But they're different in that everyone is a member of exactly one cluster--and recall that if you're not a member of a category, end of story!
We also wanted to make sure that we really did find four
clusters. So we tested the data again, with an completely different
method of finding clusters. Once again, we found four clusters. (If
it walks like a four-cluster solution and it talks like a four-cluster
solution--it must be a four-cluster solution!)
We described each cluster using the types of self-injury reported by a majority of its members. Please click here for these descriptions. Because the descriptions contain information on specific methods of self-injury, please make sure you are safe before and after reading them.
If you don't want to read a description of the clusters, you may skip to a discussion of the results or return to the previous page.
Continue to a discussion of the results
Return to background section (previous page)