This research investigates alternative methods of clustering jobs so t
hat the resultant clusters may fulfill multiple personnel functions. A
Job Activities Inventory containing three types of job descriptors (t
asks, behaviors, and abilities) was developed, and judgments on job ac
tivities and requirements were collected for 75 Navy entry-level ratin
gs (jobs). Job-level component scores were derived to form the basis f
or grouping the jobs into descriptor-specific families. Each job's clu
ster membership was then cross-tabulated across the different cluster
analysis solutions to form ''tree diagrams.'' These job family trees r
epresent a job's cluster membership across a number of cluster solutio
ns, and they organize and present job characteristic data in ways diff
erent from traditional occupational analyses. Several applications of
multibasis job characteristic information are discussed.