Construction and validation of the Subjective Underemployment Scales (SUS)

- A pool of items was generated and revised with two samples of working adults.
- The 37-item Subjective Underemployment Scales (SUS) were finalized.
- We tested correlational, single, higher order, and bifactor models.
- The SUS correlated expectedly with validity variables.
- The SUS correlated negatively with well-being variables.

Underemployment refers to work that is inferior to certain standards, such as sufficient pay or adequate use of people's abilities. Current measurement of underemployment has a number of problems, including categorization of continuous variables, lack of conceptual clarity, and overreliance on objective and single-item scales. To address these issues, the goal of the current research was to develop and refine scales assessing the six major domains underemployment identified by Feldman (1996): pay, status, field, hours, involuntary temporary work, and poverty wage employment. In two studies with different samples of diverse, working adults, we provide evidence for the internal consistency, factor structure, and the validity of the Subjective Underemployment Scales (SUS). The SUS demonstrated convergent validity by relating to similar constructs, work-related outcomes, and job attitudes. In addition, we tested different factor structures and incremental validity for the six dimensions, along with overqualification, and we found evidence for the use of a total SUS score and subscale SUS scores. The current research represents improvements in underemployment measurement and continued conceptual clarification in this area. Limitations, implications, and directions for future research are discussed.